The Linux cluster provides computational resources for BC faculty members and their research groups. This page contains information including links on how to get an account on the cluster, and how to use the cluster.
The use of the cluster is guided by the Cluster Policy Committee. The members of this committee are:
- Stefano Anzellotti, Chair (Psychology)
- Lucas Bao (Chemistry)
- Kit Baum (Economics)
- David Broido (Physics)
- Michelle Meyer (Biology)
- Sam Ransbotham (CSOM)
- Jean-Baptiste Tristan (Computer Science)
If you have comments on cluster policy, please contact firstname.lastname@example.org or one of the committee members.
Summary of Research Projects
We ask each research group to submit a short abstract describing their work using the Linux Cluster. The current projects using the cluster, arranged alphabetically, are:
Nadia Abuelezam (Nursing)
Our research focuses on the use of novel data streams to better inform epidemiology and better public health practice. We use simulation models to better understand the impact of HIV treatment and prevention interventions in Sub-Saharan Africa (South Africa and Botswana). We are also using large data sets from social networking websites to better understand sexual behavior in the United States.
Emrah Altindis (Biology)
The Altindis Lab focuses on two different projects: (i) determining the role of viral hormones in host-pathogen interactions and human disease and (ii) investigating the role of gut microbiota in Type 1 Diabetes and celiac disease. We will use this cluster for both of these projects.
James Anderson (Economics)
The project is going to econometrically estimate implicit trade costs for some 450 commodities for over 100 countries using bilateral trade flow data (around 10,000 observations). Stage 1 is this estimation. Stage 2 is the use of these trade costs to calculate their incidence on the seller and buyer sides of the market. For each commodity, the incidence calculation requires the solution of some 100 nonlinear equations in as many unknowns. In order to create confidence intervals for the incidence calculations, the nonlinear system must be solved many times.
S. Anukriti (Economics)
My research focuses on topics in development economics, demography, and the economics of gender. In ongoing work, I use data from large-scale household surveys to examine (a) whether financial incentives can simultaneously decrease fertility and the sex ratio at birth in societies with a strong preference for sons and (b) the effects of trade liberalizationon fertility and child survival in India
Amittai Aviram (Computer Science)
How Language Signals Cognitive State. The cognitive state of a speaker consists of that speaker’s beliefs about what he or she is saying — e.g., whether it is true or not and whether the speaker is certain or not — the speaker’s sentiments (feelings, judgments) about the content of the utterance, and the speaker’s internal model of the addressee’s beliefs and sentiments. We will apply deep learning techniques to existing corpora (data sets) to predict cognitive states based on clues within the utterances, which will help us to understand better how people express their cognitive states through subtle language feature.
Pierluigi Balduzzi (Finance)
The project examines whether and how time-variation in the daily return co-movements between developed markets and emerging markets as a whole can be linked to measures of developed stock markets uncertainty. Preliminary findings suggest that developed markets uncertainty has important cross-market pricing influences. The project will examine the dynamics of correlations cross equity markets using DCC-GARCH models and Regime-Shifting models with time varying Markov transition probabilities. The project will also investigate unconditional portfolio performance of policies that take into account time-varying correlations.
Lucas Bao (Chemistry)
We apply quantum chemistry methods, kinetics theories, and dynamics simulations to study the problems that are related to energy, catalysis, environment, and sustainability. Our research goal is to apply and develop new theoretical methods to understand the underlying physical principles of atmospheric reactions, electrocatalysis, photocatalysis, and materials chemistry. We focus on predicting and understanding the kinetics and dynamical behaviors of the complex chemical systems based on first-principle computations. The theoretical methods that we use include but not limited to (a) electronic-structure methods: multireference wavefunction theory, density-functional theory, composite correlated wavefunction theory, etc.; (b) kinetics and dynamics: variational transition state theory, semiclassical quantum tunneling, ab-initio molecular dynamics, etc.
Simcha Barkai (Carroll School of Management)
The aim of this project is to understand the role of higher education in shaping attitudes towards economic policy. The project will provide two types of analysis. The first is analysis of new survey data collected from Boston College students and Alumni. The second focuses on observational data, such as data available through online profiles of individuals.
Christopher Baum (Economics)
This project evaluates various behavioral risk factors in a public health context, including the effect of tobacco taxes and smoke-free legislation on mothers' smoking and babies' birthweight; the relation between smoke-free legislation and the incidence of childhood asthma; and the use of tobacco products by adolescents.
Kevin Bedell (Physics)
Our project is focused on the study of exotic collective modes in the magnetically ordered systems based on the Landau Fermi liquid theory. We use the Landau kinetic equation in the spin channel, to study the dynamics of the fluctuation to the ground state we start with. By solving the kinetic equation in the hydrodynamic region, we can determine the dispersion of the collective modes. We also calculate the spin response function of the system to further study the dispersion and the effect of the collective modes to the ground state.
Mark Behn (Earth and Environmental Science)
Our research involves developing numerical models to study active tectonic and magmatic processes in marine, terrestrial, and polar environments. Deformation and mass transport depend critically on the rheologic properties (i.e., strength) of the crust and mantle. Thus, any quantitative study of active tectonics requires a thorough understanding of the Earth’s rheology. My research group develops numerical models to relate laboratory-based rheologic and petrologic models to the large-scale behavior of the Earth. We use finite-difference and finite-element based computational approaches in two- and three-dimensional simulations. Our models are applied to a range of problems, including faulting, mantle convection, and melting and melt migration in the Earth’s mantle, as well as to societally-relevant issues, such as the dynamic response of ice sheets to climate change, global geochemical cycling, and hazards associated with earthquakes and volcanic eruptions.
José Bento (Computer Science)
We will be working on distributed and parallel optimizations algorithms and their applications to machine learning and biology. The framework will be based on the popular alternating direction method of multipliers and their variants. We will use tools like spark, OpenMP and OpenMPI to develop these tools. In addition, we will develop of series of theoretical results to improve the robustness and accuracy of these algorithms.
Theresa Betancourt (Social Work)
The purpose of this study is to examine social, structural and demographic factors interacting with intimate partner violence (IPV) to define mental, nutritional and reproductive health outcomes for women in humanitarian or conflict-affected settings in Africa. To evaluate the impact of these problems, this study will examine the prevalence, interactions and mediating effect of IPV on maternal and child well-being using data from the Demographic Health Survey (DHS) of selected sub-Saharan African countries in conflict situations.
Vincent Bogousslavsky (CSOM)
I study how measures of stock market liquidity vary over the trading day. The close of the stock market concentrates a much larger fractionof daily trading volume nowadays than twenty years ago. This change seems in part due to institutions that trade near the close because of indexing constraints. As a result, intraday liquidity patterns have changed over time. This project analyses a large data set of high-frequency liquidity measures to understand what drives variation in liquidity both over time and in the cross-section. A better understanding of stock market liquidity has important implications for market efficiency and regulation.
Henry Braun (Lynch School)
This project uses a conditional growth chart method to track student academic growth. This method is based on a regression model called quantile regression. The specific part that requires intensive computing is a simulation extrapolation (SIMEX) method that we apply to quantile regression to correct for measurement error-induced bias, since student academic achievements are measured by test scores which contain considerable amount of measurement error. The basic idea of the SIMEX method is to add simulated additional measurement error with increasing variance to the original data in a resampling-like stage, identify a trend of measurement error-induced bias versus the variance of the added measurement error, and extrapolate the trend back to the pointwith no measurement error.
Jeffrey Breeding Allison (Mathematics)
In my current research, I am studying various modularity conjectures that relate certain types of modular forms to abelian varieties and to Galois representations. These conjectures provide us with deep insight into the structure of number fields. In particular, I study spaces of paramodular forms and attempt to verify the Paramodular Conjecture due to A. Brumer and K. Kramer. The Paramodular Conjecture relates weight two paramodular forms of level N that are not in the span of Gritsenko lifts to certain abelian surfaces of conductor N. It is a degree two analogue of the Taniyama-Shimura Conjecture, which relates elliptic curves to classical modular forms. My projects include computing spaces of paramodular forms using either Fourier-Jacobi expansions or Borcherds products and computing spaces of orthogonal modular forms to investigate other modularity conjectures.
David Broido (Physics)
Our research group focusses on the study of heat transport in bulk and nanostructured semiconductor materials. Our goal is to develop an accurate theoretical approach that will allow us to provide guidance to experimental groups who perform measurements of these materials, as well as contributing to the development of new nanomaterials engineered for specific applications. We are employing state-of-the-art computational methods (for example, ab initio and adiabatic bond charge calculations of phonon dispersions, iterative solution of phonon Boltzmann equation) in this effort that require multiple fast cpu's and substantial memory.
K.S. Burch (Physics)
We look at the lattice vibrations in various materials, with a goal of understanding their role in material properties. Our interest is in comparing theoretical predictions with experimental measurements made with Infrared and Raman spectroscopy to gain insight into the origins of anomalous phonon response. The also enables us to guide future design of materials with optimized properties.
Jeffrey A. Byers (Chemistry)
The Byers group will use the Cluster for the molecular modeling of transition metal complexes. These complexes will be used as catalysts for important chemical transformations in organic chemistry, inorganic chemistry, and materials science. Theoretical investigations carried out with Gaussian and/or Jaguar will be invaluable to the catalyst development process
Rocio Calvo (Social Work)
Few studies have focused on the social determinants of health disparities among second-generation immigrants. This project explores individual-level determinants of health disparities such as race, income, social and human capital on health disparities among a
representative sample of post-1965 immigrants in the United States. We seek to answer the following questions: (1) how self-reported depression, obesity, well-being, etc. vary across ethnic/racial groups and over time as respondents enter young adulthood; (2) what individual and contextual factors determine depression, obesity, well-being, etc.; (3) what social mechanisms underlie the relationship between individual and contextual-level characteristics and the social determinants of health among the second-generation in the US.
Ryan Chahrour (Economics)
Economists have traditionally viewed prices as extremely efficient mechanisms for transmitting individuals' private information to others. My research seeks to understand when this view is justified and when and how prices may fail to fully transmit relevant information. Solving dynamic models where prices transmit information is computationally challenging and requires both new algorithms and substantial computational resources. I will use the cluster to advance my research and solve more complex versions of the model described in www.chahrour.net/Intersectoral_Linkages_Information.pdf
Xiao Chen (Physics)
My research focuses on non-equilibrium quantum dynamics. I use anumerical simulation to (1) better understand the quantum dynamics experiments in Nitrogen-vacancy center and cold atom systems, (2) A study the thermalization and entanglement dynamics in isolated many-body quantum system with or without conservation law, (3) find possible new phases in non-equilibrium dynamics and (4) design many-body quantum circuit models with possible applications in quantum computation.
Mengyao Cheng (CSOM)
This project will match firms' political contributions and spending with firms' CEO personal political contributions and spending, as well as firm characteristics such as performance. We are interested in examining the effect of the political environment on firms' disclosure choices.
Thomas J. Chemmanur (Finance)
We study that social network has first-order effect on the portfolio holdings and trades for mutual fund managers. We examine whether there is valuable information transmitted through the network or the effect is due to the herding behavior of professional money managers.
John P. Christianson (Psychology)
Numerous psychiatric conditions – including autism spectrum disorder – are characterized by abnormalities in social cognition. Thus, the description and quantification of social behavior in laboratory rodents is central to basic and translational research. Conventional ethological approaches to social behavior are fraught with challenges including bias, significant human effort and temporal accuracy. Machine learning can be applied to laboratory tests of social decision making to remedy these challenges. We will use supervised machine learning to train a convolutional neural network to identify points on interest on freely moving, unmarked rodents during various social behavior paradigms. The resulting model will then be used to produce behavioral tracking with reduced bias, higher efficiency, and increased temporal precision.
Julia Chuang (Sociology)
Land has become an explosive issue in China. The Chinese government now draws the majority of its revenues from land transactions, and lost land-use rights is now the leading cause of protest among farmers. In this inquiry, we ask whether recent policy changes, in particular a policy allowing the exchange of use rights to collective land, previously forbidden under socialist ownership rules, empower farmers or create avenues for further local government expropriation. To this end, we use text classification using word embeddings and deep learning to determine which local governments have adopted this policy. We also use a unique dataset of land grievances in China, based on evidence posted on Sina Weibo microblogs (Chinese tweets) of collective action around land grievances. This research should contribute to the literature a broad causal scheme for why the same policies benefit rural people in some localities and facilitate local state predation in others
Ali Cirik (Marketing)
This project supplements in-class coding sessions and introduces cluster computing concepts to students. The main focus is to give students another perspective and show them computing resources available beyond their computers. I will explain cluster computers in general, emphasize responsible use, and then show students how they can operate with their accounts to deal with big data, build and experiment on machine learning models, and eventually get results and make decisions. Some of the methods I will introduce are Natural Language Processing (NLP) and Generative Adversarial Networks (GAN), while the experience will be tailored for each class and student interest. We will explore and use academic-focused, publicly available data sets such as Yelp Open Dataset, Twitter API for Academic Research, Open Data on Amazon Web Services (AWS), and Google Cloud Platform (GCP) Datasets.
Peter Clote (Biology)
Our work involves developing new algorithms concerning RNA structure prediction (protein as well). Our algorithms run in times of O(n4) and O(n5) with space an order of magnitude less.
Lucas Coffman (Economics)
We conduct an at-scale randomized control trial evaluation of interventions that present accurate, clear information – largely through in-class videos – on the potential benefits and costs of schooling to almost 200,000 7th–12th grade students in the Dominican Republic. Merging with decades of administrative data as well as dozens of self-administered surveys -- done online, inclass, at home, and over the phone -- we can estimate the effects on dropouts, test scores, and matriculation to university as well as important heterogeneity of those effects.
Timothy Connolly (Biology)
The primary objective of the project is to introduce students to some basic concepts of bioinformatics and re-enforce concepts in genetics and genomics in order to improve their understanding of the biological predispositions to disease, disease processes and potential avenues to treat disease. The field of genomics and bioinformatics is having a significant impact in the field of Biology and Medicine.
Using a web user interface, and a series of real public genomic datasets, undergraduate students can bioinformatics methods used to analyze genome-scale data in regard to real world, unmet scientific challenges where genomic technologies are having an impact. A series of presentations, manuals and guide workflows will help student ability to acquire large genomic datasets and use different computational methods used to analyze these datasets. Using a guided series of small objective exercises to introduce individual approaches, students can gain hands-on experience with various approaches to genomic analyses. In doing so, students can gauge their interest in the field of bioinformatics and computational biology.
Donald Cox (Economics)
My research project "Height, attractiveness and social status among adolescents" will use the National Longitudinal Study of Adolescent Health (Add Health). Longstanding and (nearly) universal height-related norms for couples arguably introduce constraints for individuals searching for partners. This research seeks to learn about how height and attractiveness impinges on the formation and duration of adolescent romantic relationships.
Thomas Crea (Social Work)
The Ebola Virus Disease (EVD) Pediatric Mental Health Project aims to understand the physical and ecological effects of EVD on the physical health and psychosocial wellbeing of children in Sierra Leone. We aim to (1) identify ecological factors (i.e., family and community acceptance, psychosocial adjustment) influencing processes of risk and resilience among EVD-infected, affected, and unaffected children, ages 10-17, as well as caregivers, over three time points; (2) identify differences in the physical health and functioning of EVD-infected, affected and unaffected children over time, looking at physical health indicators and stress biomarkers from retinal scans; and (3) strengthen the capacity of key organizations to conduct ongoing research on vulnerable children and families’ psychosocial well-being and physical health.
Jeffrey DaCosta (Biology)
Evolutionary biology is currently undergoing tremendous growth with the emergence of next-generation sequencing technology and associated analyses, which provides unprecedented power to study evolution on a genomic scale, determine the genetic underpinnings of phenotypes, and examine those phenotypes in a phylogenetic framework. My research uses these tools to reconstruct the evolutionary history of species and populations (mostly birds), and advance our understanding of the generation and maintenance of biodiversity. My expertise in this field is transferred to undergraduates through advanced experience research labs and independent projects in which students gain exposure to the process of collecting and analyzing genome-scale data. These
skills are general to modern genomics biology, and could also be useful for students seeking careers in fields such as evolutionary biology, conservation genetics, biotechnology, or biomedical research.
Francesco D'Acunto (Finance)
We document that, since 2011, mortgage lenders reduced credit to middle-class households by 15% and increased credit to wealthy households by 21%. Credit to low-income households was unaffected. Results hold at the individual-loan level and zip-code
level, and at the intensive margin and extensive margin. The redistribution increased monotonically with the size of the lender. The collapse of the private-label securitization market, banks' risk-management concerns, wealth polarization, post-crisis policies of GSEs, or pre-crisis indebtment are unlikely to explain the results. The results appear consistent with large banks reacting more to the increased costs of origination imposed by financial regulation. Working on the cluster will allow us to run quantile-regression models that enable quantifying the effects of the redistribution of mortgage credit on households' overall wealth based on their position in the US income distribution.
Patricia Doherty (Institute for Scientific Research)
Our research efforts have included the study of ionospheric data from probes measuring electron and ion densities, auroral electron precipitation and atmospheric infrared visible and ultraviolet emissions. We develop, validate and update various models that propagate electromagnetic waves under quiescent and disturbed conditions. These models facilitate data comparisons and theoretical calculations requiring a background atmosphere, as well as providing convenient engineering solutions.
Marek Domin (Mass Spectometry)
The Linux computational cluster is used for processing large batches of data, for both targeted and non-targeted metabolomics analysis. Typical MS data processing workflow comprises, raw data file import, filtering/smoothing, peak picking, peak list deisotoping, alignment, gap filling and normalization. We use MZmine 2, an open source software toolbox for LC-MS data processing. The MZmine 2 modules cover all these workflow stages and also include additional functionality for the visualization and interpretation of the results.
Shaun Dougherty (Lynch School of Education & Human Development)
In this project, we use publicly available data on the expansion of charter schools and other schools of choice, to understand dynamic changes in the enrollment of Catholic schools in the United States. In addition, we estimate the impact of the proliferation of new schools of choice in proximity to Catholic schools to predict declines in enrollment and the conditions under which Catholic schools close. Using over two decades of nationally representative school survey on student enrollment and educator characteristics, as well as local school data in a sampling of metro areas, we plan to show whether and how the proliferation of school choice options changed the characteristics and numbers of Catholic schools.
Jan Engelbrecht (Physics)
Our research explores emergent phenomena in both physical and biological contexts where many simple interacting "entities" develop novel collective behaviour not found in the original building blocks. Our work in physics considers correlated electrons cooperating to exhibit strange behaviour in high-temperature superconductors. We have a new program in neuroscience where we extend some of the ideas of emergence in physics to consider how populations of `interacting' neurons develop collective behaviour that performs function. Specifically we consider how the dynamics of the development of synchrony in neural spike times can realize algorithms for sensory pattern recognition and binding.
Maksym Fedorchuk (Mathematics)
I plan to use the account to run computations in SAGE and Python with the goal of understanding the geometry of moduli spaces of algebraic curves. My main interest is in computing cones of nef and effective divisors on these moduli spaces, with the eventual goal of resolving several long standing conjectures regarding these cones.
Our project focuses on developing a Markov-chain Monte Carlo algorithm aimed at determining the magnetic ground state and the temperature-dependent magnetic phases of the alpha-type antiferromagnet EuCd2P2.
Hanno Foerster (Economics)
This project studies questions in labor and family economics. I develop and estimate dynamic economic models (e.g., life-cycle models or search and matching models) to study questions of high policy relevance. This involves working with data from various sources, including survey data as well as administrative data, such as Danish data from social security and tax records or German social security data. As part of this project I study the optimal design of child support and alimony policies. I also investigate what led to the abandonment of U.S. laws that regulated women’s work hours and occupational choices until the 1960s.
Vyacheslav Fos (CSOM)
Our project tries to examine whether early-life experience, military experience in particular, form future executives' traits and affect their career consequences in the firms. We look at corporate decisions, such as CEO compensation and turnover, for executives with and without military experience. We also study whether option exercise strategies differs across the so called "military executives" and their counterparts.
Solomon Friedberg (Mathematics)
Automorphic forms are fundamental in modern number theory and representation theory, and are closely related to the spectral decomposition of certain homogeneous spaces. It is often challenging to compute this spectrum explicitly, but desirable to do so, as there are many distributional questions one can ask once one has the data. The goal of this project is to compute approximations to Maass forms attached to number fields and then to study them numerically. The initial goal is the computation of Bianchi modular forms. These objects are attached to congruence subgroups of the group of two by two determinant one matrices with entries in the ring of integers of a fixed imaginary quadratic field.
Vincent Fusaru (Social Work)
This project examines the effectiveness of eviction moratoria during the COVID-19 pandemic. Drawing data from a variety of sources and multiple levels of geography, we estimate models of the number of eviction filings at the Census tract level as a function of eviction moratoria policies, local economic and public health conditions, and population characteristics.
Jianmin Gao (Chemistry)
The Gao group in the Chemistry Department of Boston College seeks to develop novel peptide inhibitors for biomolecules previously deemed undruggable. In comparison to the traditional small molecules used in medicinal chemistry, peptides harbor larger molecular footprint and hence are able to effectively inhibit protein-protein and protein-nucleic acid interactions, which often display larger contact surfaces. The Gao group develops various bioconjugation chemistries and applies them to construct peptide libraries that incorporate diverse non-natural structure motifs. Screening of such peptide libraries has proven to be a powerful way to identify initial hits for inhibiting a biomolecule of interest. The initial hits typical display suboptimal potency and specificity, which calls for structural optimization. We propose to use computational modeling, including docking and molecular dynamic simulations, to elucidate the binding mode of the peptide hits. The structural models will be used to generate hypothesis to streamline the structural optimization of the peptide inhibitors.
Glenn Gaudette (Engineering)
Our work focuses primarily on developing plant-based, edible scaffolds for sustainable production of cultured meat. We also use the same scaffolds and cell growth strategies for regenerative medicine applications. Using whole-transcriptome RNA-seq data, we aim to examine how cells respond to these scaffolds, comparing them to cells isolated from native tissue.
Alyssa Goldman (Sociology)
This project examines how social relationships and neighborhoods shape older adults' health. We explore how different neighborhood exposures and measures of social integration intersect to influence well-being in later life. We aim to better understand the mechanisms through which the social context plays a role in physical, functional, psychological, and physiological health.
Siddhartan Govindasamy (Engineering)
Future generations of wireless and mobile communication networks are expected to use a very large number of antennas to help serve a large number of mobile users. These systems, are referred to as Massive, Multiple-Input Multiple-Output (MIMO) wireless communications. System-wide performance of massive MIMO networks can be understood through a combination of analytical and simulation-based methods. In this work, we plan to simulate wireless networks with a very large number of antennas to improve our understanding of the tradeoffs between the increased complexity of such systems, and their ability to provide service to a large number of users.
Michael Graf (Physics)
Our group focus is on measurements of strongly correlated electron and magnetic systems at low temperatures. These works encompass materials that are at the forefront of modern condensed matter physics, and involve exotic low temperature phases and are a result of complex many-body interactions. One prominent technique in our research is Muon Spin Resonance (MuSR or μSR), used to probe the local distribution of magnetic fields. As a part of this technique, numerical fitting routines must be performed on the data to extract parameters based on analytical models. At present we are not able to utilize the software to its full potential based on the limited computational speed of desktop machines. The Scorpio cluster will be used to implement the more powerful and thorough routines of the software and thus extract much more detailed information than currently obtainable.
Matthew Grimes (Mathematics)
My research focuses on the geometry of moduli spaces and understanding transitions called wall crossings. I will use the cluster to perform calculations for better understanding of GIT stability. These computations will advance our understanding of curves and the
relationship between various moduli spaces, as well as address many open questions in the field.
Rob Gross (Mathematics)
Our research group is investigating new algorithms to find efficientlattice packings in moderately high dimensions (20<n<99). We hope to improve on some known bounds by using parallel methods.
Michael Grubb (Economics)
My research focuses on topics in behavioral industrial organization. In ongoing work I use historical cellular phone and electricity billing data to investigate the potential effect of bill-shock regulation on consumers. Bill-shock regulation seeks to inform consumers by requiring firms to alert consumers when high usage triggers an increase in marginal price. Importantly, the work seeks to predict how firms would change prices in response to such regulation and takes these predicted price changes into account when evaluating the policy’s impact on consumers.
The Gubbels lab studies the pathogenic mechanisms of apicomplexan parasites at the cell biological and genomic level. Next generation sequencing (NGS) and imaging are critical technologies in our tool box. We aim to build comprehensive networks of the biological processes in Toxoplasma gondii and Sarcocystis neurona by applying a variety of computational tools and mathematical models.
Pablo Guerron (Economics)
My research involves the development of fast and efficient methods to solve and estimate macroeconomic models. I am particularly interested in models that display high nonlinearities arising from borrowing constraints, default, and nonnormal shocks. The GPU cluster will be used to develop these methods. Equally important, the cluster will help to introduce students to modern computational techniques such as parallelization and programming, for example, in CUDA.
Fredrik Haeffner (Chemistry)
I collaborate with several research groups at the Department of Chemistry, where I use quantum chemistry in an effort to better understand chemical reactivity and selectivity of a variety of organic and organometallic catalyzed transformations. Examples of past and current research involve olefin metathesis, N-heterocyclic carbene catalyzed addition and substitution reactions, Pt-catalyzed enantioselective diboration of monosubstituted alkenes, chemical and spectroscopic properties of azaborines. Other interests of mine are enzyme mimics and in-silico design and optimization of catalysts for a variety of reactions.
Cal J. Halvorsen (Social Work)
My research focuses on the predictors, experiences, and outcomes of longer working lives and, in particular, self-employment in later life. In conducting this research, I use nationally representative and longitudinal datasets, such as the Health and Retirement Study.
Michael Hartney (Political Science)
This project examines the factors that influence citizens to turn out to vote in local school board elections, with a special focus on the political behavior of a school district's employees. Linking millions of voters to their school district of residence and further linking public school employees to state voter files allows me to understand the contextual factors that predispose voters to participate in local elections. In particular, I examine whether school employees are more or less likely to vote when they live and work in the same school district. Additionally, I analyze whether school employees are differentially driven to participate in response to policy decisions that their employer school district makes that have a direct influence on their occupation (collective bargaining contracts, salaries).
Joshua Hartshorne ( Psychology)
Our research is focused on learning the structure of the syntax and semantics of English verbs, specifically in verb argument structure and verb classes, using data from both existing lexical resources and large web-based experiments. We take a computational approach to analysis, using non-parametric Bayesian models and artificial neural networks.
Summer Hawkins (Social Work)
This program of research will utilize routinely-collected data from the birth certificate to examine the impact of state-level policies and economic conditions on maternal smoking during pregnancy. The National birth files contain information on every birth in 29 states from 2000 through 2009 for a total of 18 million births. For our first project we will examine the impact of cigarette taxes and smoke-free legislation on maternal smoking during pregnancy and test whether these relationships vary across different subgroups of the population. For our second project we will examine the effect of economic conditions, including recessions, on maternal smoking behaviors and determine whether the economic climate impacts mothers differently across racial/ethnic and socioeconomic groups.
Rawley Heimer (CSOM)
How do persistent cash flow shocks affect debt repayment across the distribution of households? Using individual data on natural gas shale royalty payments matched with credit bureau data for 215,639 consumers, we estimate that individuals repay 33 cents of debt per dollar of windfall, and that initially-subprime individuals repay approximately 5 times more debt than initially-prime individuals do. This difference in debt repayment is driven by changes to revolving debt balances. Finally, we show that debt repayment precedes durable goods consumption, particularly for households who were initially financially constrained. These results shed new light on how deleveraging affects household consumption.
Jamie Henzy (Biology)
The massive increase in biological sequencing data is akin to the invention of the electron microscope, in that it has the potential to provide a much finer resolution view into the underpinnings of many biological phenomena. Our project is not centered on one topic, but seeks to explore what types of new questions can be asked with this "new instrument" of biological big data. In particular, we are interested in applying comparative and computational methods to large data sets to find signals of biological significance that would remain hidden to "lower resolution" methods.
Stefan Hoderlein (Economics)
A new estimator based on Radon inverse is suggested to estimate treatment effect parameters in a triangular treatment effect model with random coefficients in the selection equation. This non-parametric estimator relies on certain smoothing parameters, which need to be chosen by running a series of simulations and computing various types of optimality criterion.
Amir H. Hoveyda (Chemistry)
The research in our group is centered on the design and synthesis of new organometallic and metal-free catalysts for practical applications in asymmetric synthesis. Although our group focuses on experimental solutions to the major problems in asymmetric catalysis, quantum-mechanical calculations (Gaussian) are a valuable aid for us in the pursuit for more efficient and selective catalytic systems. Theoretical analysis of the catalysts developed in our group helps us understand their reactivity profile and identify promising structures for future experimental studies. Our current theoretical studies involve investigation of the reactive intermediates in the catalytic cycles of the chiral Ru-based and Mo-based metathesis catalysts. The insight gained from our experimental and computational studies is employed towards design of new metathesis catalysts. Future investigations will include other methods under development in our group, for example, asymmetric conjugate additions and asymmetric ketone and imine alkylations.
David Hughes (Economics)
My work involves the development of new estimators to answer economic problems, as well as theoretical and data-based exploration of the properties of these estimators. Examples included methods for bias correction in models that allow for unobserved heterogeneity, and the application of nonparametric and machine learning techniques to instrumental variables problems. This typically involves large-scale simulations of new estimators under different data generating processes.
Megan Hunter (CSOM)
My projects involve exploring the pros/cons of firms making it harder to consumers to access information. In particularly I am working with a business to business firm. I will be exploring their clickstream data to see how their customers react when they hit an "identity wall." Certain technical papers by the firm require the consumer to fill out more information about themselves before they can access it.Material
Amy Hutton (Finance)
We are examining the relation between firm's idiosyncratic risk and the level of transparency in their financial reports. We expect to demonstrate that greater transparency leads to greater information flow and thereby greater idiosyncratic risk. The ultimate question we hope to address is whether greater transparency results in less mis-pricing of firms' traded shares.
Peter Ireland (Economics)
We are constructing and analyzing a set of dynamic, stochastic, general equilibrium macroeconomic models in which heterogeneous agents possess imperfect information either about the true structure of the economy or the set of shocks impacting on the economy. As these models are computationally intensive, our efforts focus partly on developing and implementing numerical procedures to solve them. From a substantive viewpoint, we are investigating how the actions of imperfectly informed agents propagate shocks through the economy and how government policies can be designed to mitigate the distortions and welfare losses that result from private agents' imperfect information.
Welkin Johnson (Biology)
Viruses have contributed to the evolution of all life, and we are only beginning to understand the magnitude of their contributions. In this project, we aim to elucidate the presence, abundance and potential activity of endogenous retroviral elements in vertebrate genomes, with a focus on primates and fish, with the goal of discovering host functions that originate from retroviral genome insertions.
Adam Jørring (CSOM)
My research project analyzes the effects of bank competition on discriminatory practices in mortgage lending. We find that mortgage lenders are significantly less likely to approve black applicants' loan applications despite facing similar credit risk. However, following the relaxation of interstate bank branching laws in the 1990s, increases in local lending competition reduced the approval differential between potential white and black borrowers by roughly one quarter.
Mohammad Ali Kadivar (Sociology)
Forty years after the Islamic Revolution in Iran, democracy in Postrevolutionary Iran in the theological state is still a topic of discussion. In this inquiry, we focus on the role of pro-government mobilizations in Postrevolutionary Iran to argue how and when government in Iran mobilize people. We focus our attention on organization and institutional procedures behind this kind of social mobilization. To this end, we gather data from government-sponsored news agencies and code their news into particular categories that represent how government in Iran mobilizes people. Since government owns a series of news agencies, we scrape them in order to find our key categories. Therefore, we limit our scope to recent years in order to get the most-updated data about the role of pro-government mobilizations in Postrevolutionary Iran.
Alan Kafka (Earth and Environmental Sciences)
Conducting a statistical analysis in Matlab, known as Cellular Seismology, on earthquake activity in the Pacific Northwest. Using the Cellular Seismology method, the assumption that past earthquakes delineate zones where future earthquakes are likely to occur is tested using the “the CS hypothesis”, i.e., testing the extent to which the assumption holds true in a given study area, in this case the PNW. A better understanding of the spatial and temporal variation in how well the CS hypothesis holds is important for earthquake forecasting and seismic hazard research. I'm using two different methods to conduct this analysis over various time spans in order to compare and contrast different seismic behaviors detected by each of the methods. In addition to a broad assessment of the region, various parameters will also be isolated in the data to assess the earthquake behavior using both methods.
Alan Kafka and John Ebel (Earth and Environmental Sciences)
Although earthquake prediction remains an elusive goal, it is possible to forecast the general characteristics of future earthquakes at some level of detail. Seismologists use the term "earthquake forecast" to refer to a statement of the long-term probability of one or more earthquakes occurring in a region. Our research on earthquake forecasting is focused on discerning the level of detail that can be known about the spatial and temporal characteristics of future earthquake processes. We are investigating the extent to which the distribution of seismicity in a region delineates where future earthquakes are likely to occur, as well as the extent to which non-random patterns in the temporal distribution of seismicity might indicate increased probability of earthquakes occurring.
Gabor Kalman (Physics)
Our research focuses on the theoretical analysis combined with computer simulation of the properties of strongly coupled plasmas, complex (dusty) plasmas in particular . Complex (dusty) plasmas consist of mesoscopic grains immersed in the background of gaseous plasma of electrons, ions and neutral atoms. Such systems occur in a variety of situations. Our aim is the study of the collective behavior in such systems: i.e. phenomena resulting from the cooperative participation of many particles. Such a collective behavior is the hallmark of strongly coupled systems. Six major areas are proposed for study: (i) magnetic interaction between grains carrying a magnetic dipole moment; (ii) propagation of waves in complex plasmas in two dimensional and three-dimensional configurations; (iii) micro-instabilities generated by beams of ions or grains penetrating into the complex plasma; (iv) stochastic and disordered behavior in complex plasmas; (v) phase transitions in complex plasmas; (vi) feasibility study for the cryogenic generation of complex plasmas on the surface of liquid helium. One of the principal approaches to be used in the proposed investigations relies on an analytic method referred to as the Quasi Localized Charge Approximation that has been successfully used previously.
Broader impacts of the research will contribute to improved understanding of problems of fundamental importance in plasma and condensed matter physics. Establishing techniques suitable for the manipulation of mesoscopic structures will also have impact on engineering applications.
Evan Kantrowitz (Chemistry)
Work in my laboratory is centered around an understanding of the relationship between protein structure and function and in particular, how protein structure relates to catalysis, metal binding, and cooperativity in enzyme systems. The BC research cluster will be used for: (1) Calculations involved in protein structure determination by X-ray crystallography. For this work will be make use of software such as CNS and XPLOR. (2) Calculations of how small molecules bind to receptor targets for drug design. For this work we have written a series of scripts that interface to a proprietary MySQL database. Software for these calculations includes AUTODOCK, DOCK5, GOLD, GLIDE and SURFLEX. (3) Molecular dynamics and molecular mechanisms calculations on our systems will be used to better understand their mechanism of action. For this work we will make use of software such as GROMACS and NAMD. For the knowledge gained from these studies we hope to develop new classes of inhibitors that can potentially become drugs from the treatment of viral infections, malaria, diabetes and cancer.
Krzysztof (Kris) Kempa (Physics)
Electromagnetic properties of nanostructures determine the behavior of sensors, solar cells and other novel devices. The research activity in the Physics Department at BC has been, in part, devoted to making and studies of nanostructures. As a part of this effort, the theory group developed advanced computer simulations of the nanostructures. These simulations involve numerical solutions of Maxwells equations, in various domains (time, space, frequency and momentum), and with realistic parameters for the materials employed. These simulations guide the experimental groups involved in studies of the nanostructures.
Elizabeth A. Kensinger (Psychology)
Our research examines the neural activity associated with memory processes. We are particularly interested in understanding how neural activity differs when information is successfully remembered versus when it is forgotten, and how the neural processes that correspond with accurate memory differ for emotionally meaningful experiences versus for more mundane ones. To examine these questions we use Matlab-based software in order to analyze the hemodynamic (blood-flow) responses throughout the brain as individuals are remembering events.
Shakeeb Khan (Economics)
The work involved will be using computer software packages such as Matlab to explore statistical properties of new inference procedures for dynamic nonlinear panel data models in econometrics, that pertains to my ongoing academic research. I will be hiring one or two PhD students to assist me on this project.
Do Yoon Kim (CSOM)
I study the different roles of corporate and open source contributors in creating economic value. Towards this, I collect massive amounts of data consisting of source code, patents, and product introductions.
Marios Kokkodis (Information Systems, CSOM)
My research focuses on understanding and optimizing design aspects of online marketplaces. I analyze massive amounts of data from multiple sources and propose technical/econometric models/algorithms that increase the transactional efficiency of these marketplaces. I will use the cluster to store, query and process data, and to train and evaluate predictive/econometric models.
Danial Lashkari (Economics)
My work typically straddles the micro and macro sides of economics, meaning that I construct models that explain the data on the behavior of firms or consumers at the micro level, as well as the aggregate behavior of the economy at the macro level. For instance, in ongoing research, I estimate a novel model of firm-level production that accounts for the role of adoption of information technology in reshaping the span and scale of firm production. The work then studies the implication of these changes at the aggregate level for macroeconomic outcomes such as the share of total income generated by firms that goes to workers (rather than the owners of firms). I intend to use the cluster for the estimation of models such as this using micro-level data and then computing the aggregate implications of the model at the macro level.
Mariana Laverde (Economics)
This paper studies the limits of school choice policies in the presence of residential sorting. Using data from the Boston Public Schools choice system, I show that white prekindergarteners are assigned to higher-achieving schools than minority students, and that cross-race school achievement gaps under choice are no lower than would be generated by a neighborhood assignment rule. To understand why choice-based assignments do not reduce gaps in school achievement, I use data on applicants' rank-order choices to estimate preferences over schools, and consider a series of counterfactual assignments. I find that half of the gap in school achievement between white and Black or Hispanic students is explained by minorities' longer travel distance to high-performing schools. Differences in demand parameters explain a smaller fraction of the gap, while algorithm rules have no effect.
Tzuo Hann Law (Economics)
I work macroeconomic models that are consistent with micro level data of varying degrees. The cluster will be used to study (a) income inequality in Germany (b) inequality and entrepreneurship in China in the face of corporatization of formerly state owned enterprises (c) finance and banking in the US and (d) the application of machine learning methods to equilibrium labor market models.
Deishin Lee (CSOM)
This research studies how supply chain structure affects customer service level and raw material utilization. We investigate the tradeoffs between the cost of labor and capital resources and the cost of under-utilizing raw material (natural) resources.
Lian Fen Lee (CSOM)
This project examines the relative accuracy of management and analyst forecasts. We predict that analysts’ information advantage resides at the macroeconomic level. They provide more accurate long-horizon earnings forecast than management when a firm’s fortunes move in concert with macroeconomic factors such as gross domestic product and energy costs. In contrast, we expect management’s information advantage to reside at the firm level. Their forecasts are more accurate than analysts when management’s actions, which affect reported earnings, are difficult to anticipate by outsiders. Examples include when the firm’s inventories are abnormally high, the firm has excess capacity, or is experiencing a loss. Last, while analysts are commonly viewed as industry specialists, it is unclear whether analysts have an information advantage over managers at the industry level. Managers are also likely to have significant industry expertise and knowledge. They need to understand industry dynamics and demand to effectively run an operating firm.
Jacqueline V. Lerner (Lynch School)
We will be conducting analyses of survey data from the Connecting Adolescents’ Beliefs and Behaviors (CABB) Study, a 3-year study with adolescents from the New England area. Our main research question is whether intentional self-regulation is the process through which adolescents who report positive virtues are able to turn them into behaviors consistent with that character; i.e., whether it helps them “do the right thing.” The study was funded by a grant to Jacqueline V. Lerner, Ph.D. and Sara K. Johnson, Ph.D. from the John Templeton Foundation.
Rebekah Levine Coley (Lynch School)
Using Add Health data on a representative sample of Americans followed through adolescence (3 waves of data covering ages 12 through 26; n = 20,745), this research will assess rich, multi-reporter measures of social norms, gender roles, and parenting processes as well as genetic polymorphisms. Data analyses will employ multilevel growth modeling as well as semiparametric mixture models and regression models. This research offers the potential to understand causes and correlates of youth risk behaviors and to identify high risk youth. Results will inform development of effective intervention programs and policies to diminish adolescent engagement in health risk behaviors.
Arthur Lewbel (Economics)
General Doubly Robust Identification and Estimation: Consider two different parametric models. Suppose one model is correctly specified, but we don't know which one (or both could be right). Both models include a common vector of parameters, in addition to other parameters that are separate to each. An estimator for the common parameter vector is called Doubly Robust (DR) if the estimator is consistent no matter which model is correct. We provide a general technique for constructing DR estimators, which we call General Doubly Robust (GDR) estimation. Our GDR estimator is a simple extension of the Generalized
Method of Moments, with analogous root-n asymptotics. We illustrate the GDR with a variety of models, including average treatment effect estimation. Our empirical application is an instrumental variables model where either one of two candidate instruments might be invalid.
Zhuoxin Li (Information Systems, CSOM)
I expect to use the cluster for running computation-intensive statistical analysis with large datasets, using software packages R, Matlab, and Stata. Ongoing projects include Bayesian estimation of large-scale consumer decision marking problems. If feasible, I may utilize parallel computing to speed up data processing and analysis.
Zhushan Li (Lynch School)
My research focuses on modeling differential item functioning (DIF) under the framework of generalized linear and nonlinear mixed models via penalized estimation. Large-scale simulation studies will be conducted to examine the power of this approach and to compare the global performance with other commonly used DIF detection procedures in terms of hit rates and false alarm rates.
Miao Liu (CSOM)
We are building state-of-the-art machine learning models to predict corporate earnings and examining if this is a superior approach to assess a firm’s fundamental value than other approaches in the literature. The ultimate goal is to understand the value of corporate financial reporting to investors.
Kathryn Lindsey (Mathematics)
This research project in pure mathematics investigates dynamical properties of certain piecewise linear self-maps of an interval. The goal of the project is prove theorems about the structure of the set of Galois conjugates of growth rates of generalized beta transformations. Numerical computation of various mathematical sets associated to this infinite family of functions provides insight into properties which we seek to formulate and prove as theorems
Nan Liu (CSOM)
This project is motivated by the use of appointment templates in healthcare scheduling practice. We study how to offer appointment slots to patients in order to maximize the utilization of provider time. We develop two models, non-sequential scheduling and sequential scheduling, to capture different types of interactions between patients and the scheduling system. The scheduler offers either a single set of appointment slots for the arriving patient to choose from, or multiple sets in sequence, respectively. This is done without knowledge of patient preference information. For the non-sequential scheduling model, we identify certain problem instances where the exhaustive policy (i.e., offering all available slots throughout) is suboptimal, but show through numerical results that for most moderate and large instances greedy performs remarkably well. For the sequential model we derive the optimal offering policy for a large class of instances, and develop an effective and simple-to-use heuristic inspired by fluid models.
Shih-Yuan Liu (Chemistry)
The research in the Liu group is focused on the development of boron(B)–nitrogen(N)- containing heterocycles, specifically azaborines, for potential applications in biomedical research and materials science. Azaborines are structures resulting from the replacement of two carbon atoms in benzene with a boron and a nitrogen atom. Azaborines closely match the size and shape of ordinary benzene rings, but most of their other physical, chemical, and spectroscopic properties are significantly altered. Computational studies will be invaluable to our efforts in understanding the electronic structure and spectroscopic features of azaborines, and the mechanism of reactions they undergo.
Qiong Ma (Physics)
Our group focuses on electron transport properties in two-dimensional vdW materials. We would like to use first principles to theoretically calculate the electron band structures, which are essential to help us understand the experimental observations.
Sean MacEvoy (Psychology)
We study human object recognition and learning using a combination of behavioral and brain imaging techniques. Our present goal is to understand the brain mechanisms supporting the recognition of multiple simultaneously-viewed objects, and how these mechanisms tolerate relationships between newly-recognized objects. Real-world objects are difficult to use in this paradigm, owing to variability between observers expertise with each object. Instead, we will generate large families of novel "nonsense" objects that must be screened by several strict criteria.
Brooke Magnus (Psychology)
Our research focuses on the development and evaluation of novel item response theory (IRT) models for analyzing multivariate survey and questionnaire data. This particular project addresses the issue of zero inflation in self-reported symptom frequencies in non-clinical samples. We have proposed a Bayesian multidimensional zero-inflated IRT model for this purpose. The cluster will be used to conduct simulation studies to empirically evaluate the performance of this model across a variety of data-generating conditions
James Mahalik (Education)
Australian males are dying earlier from an array of preventable diseases (e.g., heart disease, suicide). One of the likely contributors is men’s consistently documented pattern of engaging in health behaviors that increase the risk of preventable disease. For example, whereas the medical field recommends such things as eating a healthy diet, limiting alcohol consumption, and engaging in preventive health care, men are less likely to enact all of these heart-healthy behaviors. This study addresses the objectives of the Ten to Men study by examining key determinants of Australian males’ health, including critical developmental periods, to identify explanations for men’s health and adoption of health risk behaviors in order to develop interventions and policy aimed to improve their health. Factor analytic techniques will be used to create composite measures for the variables of interest
Alan Marcus (Finance)
Our project aims to identify the factors that affect explanatory power in regressions of stock return time series against several market indexes. In general, market models show poor performance in explaining variation in stock returns, resulting in surprisingly low R-squares. However, the variation in R-square itself is considerable, a result that may indicate that R-square may be a proxy for other risk factors, reflect firm-specific information, or simply reflect noise. Using all the U.S. stock returns available since the 1930s we try to infer the determinants of R-square and their effects on asset prices.
Michael O. Martin and Ina V.S. Mullis (Education)
The TIMSS & PIRLS International Study Center at the Lynch School conducts two major ongoing programs of international assessment of student achievement. TIMSS (Trends in International Mathematics and Science Study) involves more than 60 countries, and has been reporting on fourth- and eighth-grade student achievement in mathematics and science on a four-year cycle since 1995. Beginning in 2001, PIRLS (Progress in International Reading Literacy Study), with more than 40 countries participating, reports fourth-grade students' reading achievement on a five-year cycle. TIMSS and PIRLS use sophisticated multiple-imputation techniques to derive student achievement measures in mathematics, science, and reading, and the resulting data require computer-intensive statistical procedures to estimate population statistics and their standard errors.
Christopher Maxwell (Economics)
This projects studies the extent to which closed-access criminal records policies impact employment opportunities of ex-offenders. The hypothesis to be tested is that in the absence of public access to criminal records, employers will statistically discriminate against applicants who demonstrate attributes similar to the perceived criminal such as ethnicity, geographic location, age, sex, socio-economic status and gaps in employment history. And to that extent, access policies will have at best a marginal impact on employment opportunities for ex-offenders. The analysis, which builds on earlier work by Finlay, will focus on the 1997 National Youth Longitudinal Survey data and look closely at the differential employment impacts of policies among states that allow for more open or closed access to criminal records.
Michael McDannald (Psychology)
This project looks to explore the effect of realistic life-stress and its connection to Major Depressive Disorder (MDD). Prior research has identified Reinforcement Learning as an observable behavior that is affected by both domains. I aim to observe and computationally model sub-process of reinforcement learning such as explore/exploit behavior and prediction error under both stressful and not-stressful conditions. Through this method, I hope to identify a mechanism in human decision making that plays a role in the onset of MDD.
Sarah McMenamin (Biology)
Our research focuses on the molecular and endocrine signals that control and coordinate vertebrate development. Using whole transcriptome RNA-seq data, we are examining gene expression within different tissues and developmental stages of zebrafish and other teleosts.
Carl McTague (Computer Science)
The goal is to develop software to efficiently compute with Hopf rings (elaborate algebraic gadgets having an addition, two products, and a coproduct). And to use it to perform intricate calculations to solve open problems in homotopy theory – for example, to determine the homotopy type of the string bordism spectrum MO<8> at the prime 3.
Michelle Meyer (Biology)
The biological roles of RNA, beyond encoding proteins, have expanded in the last decade to include a diversity of important gene regulatory functions in nearly all living things. At the same time, genome sequencing efforts have produced a wealth of data that can be mined tostudy the evolution of non-coding RNAs (ncRNAs), as well as identify previously unknown non-coding RNAs. We are particularly interested in RNA structures that bind proteins to control gene expression. The predominant methodology for the discovery of ncRNAs is comparative genomics. Using the massive amounts of sequence information generated by microbial sequencing projects, and various metagenomic projects(such as the Human Microbiome Project) we apply a variety of computational tools to discover new structured mRNA elements that are hypothesized to control gene expression. In particular, we use RNA structure alignment searches using programs built on stochastic context free grammars.
Robert Meyerhoff (Mathematics)
This project enumerates and analyzes families of one-cusped complete hyperbolic 3-manifolds with low-area cusp neighborhoods. We build upon computational techniques that were used to analyze small-volume compact hyperbolic 3-manifolds and to establish the Weeks manifold as the minimum-volume compact hyperbolic 3-manifold. Advanced precision and error-bound techniques help provide for empirical results and rigorous results. We intend to use these results, the theory of Dehn fillings, and the theory of Mom technology to analyze certain classes of low-volume one-cusped hyperbolic 3-manifolds.
David Miele (Lynch School)
This project aims to estimate the magnitude and significance of the associations between a number of psychological and educational constructs, including critical consciousness, motivation, well-being, and academic achievement.
Igor Minevich (Mathematics)
I am working with seven undergraduates in exploring the popular Lights Out Puzzle. The standard puzzle consists of a rectangular grid of squares that can be light or dark. Clicking a square makes it and the squares that are horizontally or vertically adjacent switch between light and dark, and the goal of the puzzle is to start with all squares dark and click just the right ones so they all become light. It is known that there is always at least one solution; we are exploring how many there are, which is an interesting and difficult problem in linear algebra over the field of two elements. We will also be working on variants of the Lights Out Puzzle that are played on a rectangular grid and exploring a connection between the puzzle and Sudoku grids. The computations involved require working with large numbers of large matrices, so both time and space are an issue.
Yi Ming (Earth and Environmental Sciences)
My research group strives to elucidate the physical mechanisms governing Earth’s climate system, and to apply the fundamental understanding to practical issues of societal and policy importance. We will perform idealized and comprehensive climate model simulations on the cluster to study how climate change may affect precipitation patterns (e.g. droughts and floods) and extreme events (e.g. hurricanes, wildfires and winter storms), a topic with profound impacts on local populations and ecosystems.
Udayan Mohanty (Chemistry)
Electrostatic interactions between charges in solution and between the charges along the backbone of RNA play a delicate role in determining its overall stability. The importance of electrostatic interactions is apparent from the fact that magnesium influences the biological activity of tRNA and Tetrahymena group I intron, for example. We are studying the nature of interaction of magnesium ions with RNA bases. The approach combines Grand Monte Carlo simulations, Poisson-Boltzmann calculation and high-level ab initio theory to determine the binding free energy for magnesium ions with RNA bases. Another goal of the project is to use Monte Carlo simulations to quantitatively determine the effects of phosphate-phosphate repulsion on DNA stiffness in vitro. Finally, we will investigate by Brownian dynamic simulations the effects on the end-to-end contact probabilities for 200-bp DNAs of binding at different degrees of saturation for HMGB proteins.
Babak Momeni (Biology)
Communities of interacting microbes are abundant in nature. They playimportant roles in ecosystems (e.g. by cycling carbon), in human health (e.g. by causing infections), and in industry (e.g. by degrading toxic waste). We build mathematical models to study how cell-level properties of members shape the overall functions of multispecies microbial communities.
Sara Moorman (Sociology)
I am using data from the National Social Life, Health, and Aging Project (NSHAP) to examine dyadic effects of marital quality on older adults' well-being, including whether there are any differences according to gender.
James P. Morken (Chemistry)
Research in the Morken group focuses on the development of new catalytic enantioselective processes and their application to natural products synthesis. Our current work in asymmetric catalysis focuses on the design and study of rhodium and palladium complexes for enantioselective allylation and dimetallation of alkenes. Computational studies (Gaussian and MacroModel) of reaction mechanisms and catalyst structures often complement experimental studies. Combined, the two approaches enhance our ability to design effective new reactions. Our most recent DFT studies on the Pd-catalyzed addition of organometallic reagents to enones, revealed an unprecedented and unexpected reaction mechanism, which is forming the basis for many of our research directions.
Julie Holland Mortimer (Economics)
Incomplete product availability is an important feature of many markets, and ignoring changes in availability may bias demand estimates. We study a new dataset from a wireless inventory system on vending machines to track product availability every four hours. The data allow us to account for product availability when estimating demand, and provide a valuable source of variation for identifying substitution patterns when products stock-out. We develop a procedure that allows for changes in product availability when availability is only observed periodically. We find significant differences in demand estimates: the corrected model predicts significantly larger impacts of stock-outs on profitability.
Alicia Munnell (Carroll School of Management)
We are currently working on a project to estimate 401(k) fees using common options in mutual funds from the Thomson Financial Ownership and the New York Stock Exchange’s Trade and Quote databases, both available through Wharton Research Database Services. These data sets are large (generally about 60GB per month of data for 5 years of data), so basic manipulations must be done to reduce the size then analyze the data.
Charles Murry (Economics)
I use the cluster to solve economic models and to combine these models with data to estimate model parameters. The models are used to make predictions about "what-if" scenarios in order to economic theory and policy. For example, what will happen to consumer welfare after a merger between American Airlines and US Airlines? I have a particular interest in models that have multiple equilibria and large-scale non-linear optimization.
Jia Niu (Chemistry)
Research in the Niu group focuses on the development of new tools for the investigation and regulation of key biochemical processes taking place on the surface of and within a cellular system. One part of the research is the design and synthesis glycomimetic synthetic polymers with unique architectures for the engineering of cell surface glycome. Theoretical investigations carried out with Gaussian or Jaguar will be invaluable for the understanding of new polymerization processes.
Anant Nyshadham (Economics)
I estimate returns to entrepreneurship in Thailand using a novel extension to panel data methods and recover estimates of structural parameters that reveal whether learning about ability drives the entrepreneurship decision in additional to the role financial constraints might play. I find large, positive average returns, but low marginal returns. Structural estimates suggest that learning about comparative advantage in entrepreneurship plays an important role in the sorting decision. Self-reported data on expected incomes support the importance of learning. This study helps reconcile confounding results from the literature. Heterogeneous ability and churning suggest that some households benefit from consulting, training, and financing interventions, while others do not and are, therefore, unlikely to adopt.
Jordan Nickerson (Finance)
Using a structural model, I examine the distortionary effects of frictions in the CEO labor market. Firms experience productivity shocks over time and either outgrow or underutilize their incumbent CEO's talent, but keep their manager to avoid a switching cost. The decision to replace a manager depends on the magnitude of the cost and dispersion of CEO talent. This results in an endogenous supply of available CEO labor which must be determined simultaneously with the optimal policy function of the firm.
Heather Olins (Biology)
As microbial ecologists, we are interested in questions relating about community structure: which microbes are present and how abundant or rare are they, how to microbial communities develop and change over time, and what environmental factors influence these communities. We are also interested in what these communities are doing and how microbial activity influences local (and global) habitats. We are currently working in local wetlands, and are particularly interested in carbon cycling.
Claudia Olivetti (Economics)
The research project focuses on intrahousehold resource allocation and estimation of bargaining power within household. I will use a large-scale consumer panel data and retail scanner data to estimate the resource share of family member within household. Then I will infer the bargaining power of the family member from the estimated resource share.
Theodore Papageorgiou (Economics)
Exploiting results from the literature on non-parametric identification, we make three methodological contributions to the empirical literature estimating the matching function, commonly used to map unemployment and vacancies into hires. First, we show how to non-parametrically identify the matching function. Second, we estimate the matching function allowing for unobserved matching efficacy, without imposing the usual independence assumption between matching efficiency and search on either side of the labor market. Third, we allow for multiple types of jobseekers and consider an “augmented” Beveridge curve that includes them. Our estimated elasticity of hires with respect to vacancies is procyclical and varies between 0.15 and 0.3. This is substantially lower than common estimates suggesting that a significant bias stems from the commonly-used independence assumption. Moreover, variation in match efficiency accounts for much of the decline in hires during the Great Recession.
Gorica D. Petrovich (Psychology)
Our research focus is on feeding behavior and the neural substrates of cognitive and hedonic influences on the motivation to eat. We study how learning and memory, palatability, stress, and novelty motivate or inhibit food seeking and consumption independently of physiological hunger in animal models. Machine learning will enable us to track, quantify, and compare multiple, specific behaviors in males and females in an unbiased manner, during food-motivated tasks.
Marcie Pitt-Catsouphes (Social Work and CSOM)
The Linux Cluster is used for work with the Generations of Talent study conducted by the Sloan Center on Aging & Work (reporting into the GSSW). The Generations of Talent Study examines the priorities and needs of employees of different ages who work in different countries. We assess employees’ past and current quality of employment and their future work-related transitions. Three key research questions are:
- Do employees’ priorities and use of workplace-based resources for quality of employment vary by age, career stage, and life stage?
- How do organizational policies, programs, and practice influence employee engagement, job satisfaction, work productivity, and career transitions in different countries?
- How does country context, such as public policies and cultural orientations, influence employees’ quality of employment?
Jeffrey Pontiff (Finance)
Most investments research uses Fama-Macbeth (1973) t-statistics to test whether or not an independent variable is related to future stock returns. Fama-Macbeth use a two-step procedure that is robust to cross-sectional correlation. Pontiff (1996) proposes an extension of Fama-Macbeth's procedure that is also robust to time-series correlation. Pontiff's method is based on modeling Fama-Macbeth slope coefficients with a moving-average process. We propose to compare the performance of Pontiff's statistic to other corrections by using simulations.
Emily Prud'hommeaux (Computer Science)
Our research focuses on developing computational algorithms to automatically identify areas of pragmatic difficulty in the discourse of young adults with autism spectrum disorder. We will also be training deep learning automatic speech recognition models to generate transcripts of speech recordings.
Ying Ran (Physics)
Our research group studies quantum condensed matter materials including frustrated quantum magnets, high temperature superconductors, quantum hall systems and topological phases of matters. Close to zero temperature, the electronic motion in these systems are described by quantum mechanics. However because of the strong interactions between
the vast number of electrons, some novel behaviors in these materials are so striking that they cannot be described in terms of the original electronic degrees of freedom; instead they are described by completely new emergent collective degrees of freedoms. The novel physics of these emergent degrees of freedoms sometimes are in analogy of the quack, lepton, and gauge fields in the high-energy physics context, but are now realized in a condensed matter compound. To study these emergent physics qualitatively and quantitatively, we use numerical methods such as functional renormalization group, variational quantum Monte Carlo, and exact diagonalization of large sparse matrices.
Bryan Ranger (Engineering)
Our research focuses on developing computational algorithms for biomedical image and signal analysis. Current projects include interpretation of ultrasound images and analysis of sensor data from portable imaging systems.
Sam Ransbotham (Information Systems, CSOM)
Our research introduces a new disaggregated formulation of the Generalized Assignment Problem. Based on the reformulation, we are able to introduce unique strong inequalities. We test the strength of this formulation on a set of standard benchmark problems and show the new formulation to be significantly stronger than other known formulations.
Tracy Regan (Economics)
We explore the impacts of pseudo-mature behaviors (PMBs) and school activities on early adult outcomes, specifically the likelihood of college attendance and annual earnings. We define the PMBs to include (binge) drinking alcohol, cigarette smoking, and sexual activity. A student's school activities involve his/her participation in sports, non-sports, and mixed clubs. The incorporation of such measures into our analysis highlights the importance of social identity, a concept that is popular in sociology but often not traditionally emphasized in economics. We use data obtained from the Restricted-Use National Longitudinal Study of Adolescent Health (AddHealth). A two-step estimation procedure is used that allows for the possible endogeneity of the PMBs. Much of the literature has struggled to find good identifying instruments, but we offer new and viable alternatives. Consistent with much of the prior research, controlling for the endogeneity of the PMB eliminates much of these variables' explanatory powers, with the exceptions being for females who smoke and males who have had sexual intercourse. While the results do differ across the genders, many of the school-level activities are statistically significant and create positive and sizable impacts on later-in-life outcomes.
Jonathan Reuter (Finance)
Using detailed over-the-counter bond transactions and holdings of mutual funds, we plan to investigate the effect of liquidity shocks to mutual funds on their holdings. We hope to better understand how the behavior of mutual funds affects the prices and liquidity of their holdings.
Maureen Ritchey (Psychology)
Research in my lab is focused on the neuroscience of human memory and emotion. We use functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) to measure brain activity, which we then relate to performance on cognitive tasks. This work involves the processing and analysis of very large datasets, which we primarily do with Matlab and R-based tools.
Ehri Ryu (Psychology)
Our research involves the development and evaluation of statistical methods for analyzing multivariate multilevel data. The particular projects include model fit assessment of multilevel structural equation models, measurement equivalence in confirmatory factor analysis framework, multilevel multi-group comparison, and statistical inference about indirect effects. The cluster will be used to conduct simulation studies to empirically evaluate the performance of statistical methods.
Erika Sabbath (Social Work)
The Boston Hospital Workers Health Study (BHWHS) is a data-sharing and intellectual partnership between the BC- and Harvard-based study team and the two large hospitals within Mass General Brigham, a large health system and Massachusetts’ largest employer. BHWHS consists of multiple longitudinal sources of employer administrative data from the two study hospitals, linked at the individual worker level with survey data on emerging and established occupational exposures and experiences.The goal of BHWHS is to improve the health of the entire hospital workforce—not just bedside patient care providers—and reduce within-workforce disparities by identifying the roots of those disparities in the conditions of work. In our study, we will test two central hypotheses: (a) that specific policies, working conditions, and exposures act as social determinants of health within the hospital work environment; and (b) that policies aimed at improving working conditions have the potential to narrow or widen occupational, racial, and wage gaps in health within the hospital workforce. The specific aims of the proposed study are: 1) to broaden the scope of our occupational health research on hospital workers by expanding BHWHS to include all workers at our two study hospitals, not just those performing bedside patient care, with an explicit focus on low-wage service workers; 2) to characterize relationships between working conditions and health disparities by worker race, occupation, and wage level in the hospital workforce; 3) to evaluate effects of organizational and public policy changes on worker health and well-being, and determine whether observed effects vary by worker race, occupation, or wage level; and 4) to inform occupational health policy and practice. The expected impact of this work is to broaden the scope of occupational health research and practice in hospitals by demonstrating the necessity of analyzing the contribution of working conditions and policies to disparities in health and well-being in the hospital workforce.
Ronnie Sadke (Finance)
This research presents a new pattern in the cross-section of expected stock returns. Stocks tend to have relatively high (or low) returns every year in the same calendar month. We recognize the annual cross-sectional autocorrelation pattern documented in Jegadeesh (1990 at lags of 12, 24, and 36 months as part of a general pattern that lasts up to 20 annual lags, superimposed on the general momentum/reversal patterns. This pattern explains an economically and statistically significant magnitude of the cross-sectional variation in average stock returns. Volume and volatility exhibit similar seasonal patterns but they do not explain the seasonality in returns. The pattern is independent of size, industry, earnings announcements, dividends, and fiscal year. The results are consistent with the existence of a persistent seasonal effect in stock returns.
Linda Salisbury (Marketing)
This project examines the impact of post-service customer satisfaction inquiries on subsequent purchase behavior. We are using the Linus cluster to examine service transaction and customer relationship management data from a North American retail service provider to test whether repeated customer satisfaction inquiries can negatively affect purchase behavior. We also examine the extent to which repeated inquiries may or may not have a synergistic effect with other relational contacts. Preliminary findings offer evidence that "over-surveying" customers can be detrimental, and the magnitude of the impact varies across customer segments.
Geoffrey Sanzenbacher (Economics)
Despite the potential influence of secondary protections on access to new treatments, generic entry and formulations, and ultimately prices, the use of these patents and exclusivities is understudied. Little is known about how these protections are used, how these uses vary across the life-cycle of a product, and whether certain exclusivities – like the powerful pediatric extension – actually serve their desired end. This series of three papers will use a comprehensive dataset of drugs that were approved by the FDA between 1984 and 2016 to: 1) describe the use of secondary protections; 2) identify varying patterns of how the protections are used across products and over a drug’s lifecycle (e.g., securing exclusivities at the end of a patent versus throughout a drug’s life); and 3) determine whether drugs receiving pediatric extensions help meet pediatric need (or simply extend the period of monopoly pricing for an additional six months).
Natalia Sarkisian (Sociology)
I am using data from the Health and Retirement Study to examine the reciprocal relationship between employment and caregiving to parents and the variations in that relationship by race and gender.
Fabio Schiantarelli (Economics)
We explore the evolution of attitudes of European immigrants to the US using the General Social Survey and, more specifically we examine how they vary across generations. We examine whether they converge(or not) to the US prevailing norm. We do this for various attitudes concerning religion, morality, gender issues, sexual mores, political orientation, etc.. We also intend to link the attitudinal data from the GSS to census data. The census data will be used to examine the evolving ethnic composition of US counties.
Juliet Schor (Sociology)
Connected consumption, characterized by other names such as “collaborative consumption” (Botsman and Rogers, 20101) and the “new sharing economy”, is based on a culture of access, use, and re-circulation of used goods as alternatives to traditional private ownership. With the potential to foster peer-to-peer learning and social connection, ecological sustainability and economic opportunity, connected consumption has the potential to transform mobility, shopping, travel, work practices, living arrangements, service provision, household production and learning. Over the course of this project, we are studying practices of connected consumption and the connected economy. Since 2011 we have studied ten cases. The first were non-profits, beginning with time banking, which is a non-profit exchange practice in which services are exchanged for units of time. The second is open learning, which is a sector of post-secondary education that consists of organizations that facilitate free or low-cost, peer-oriented, accessible learning experiences. The third is a local food swap. In 2013 and 2014 we added a case study of a makerspace, and case studies on Airbnb, RelayRides and TaskRabbit. In 2015 we began studies of delivery services and ridesourcing providers. We have also constructed a large nationwide database of Airbnb listings. Our intent is to explore the sociological significance of connected consumption, specifically focusing engagement, expertise, and efficacy, as well as how existing structures of inequality are reproduced or broken down in these initiatives.
Lawrence T. Scott (Chemistry)
The Linux computational cluster is used by our research group primarily for the study of geodesic polycyclic aromatic hydrocarbons (PAHs) and synthetic intermediates used in their preparation. Using the Gaussian modeling software, we perform high level density functional theory calculations to optimize the geometries of these molecules and accurately predict their properties, including NMR and UV-vis absorption spectra.
Arvind Sharma (Woods College Faculty)
We are studying the impact of minimum wages on intergenerational income mobility in the US by creating a unique panel data of Metropolitan Statistical Areas (MSAs) from 1998 to 2016 using publicly available economic datasets, and Raj Chetty’s quintile transition matrices of child and parent income.
Ce Shen (Social Work)
Our research projects focus on cross-country analysis. A special emphasis is given to the study of people's values and beliefs over time. In ongoing work, we use data from large scale global surveys, such as World Values Survey and barometer surveys from several countries.
Ewa Sletten (Accounting)
We compare earnings' informativeness in bad-news and good-news quarters. We predict that when the news reaching the market during a quarter is negative, a greater proportion of it is released through earnings than when the news is positive, because of various factors that prompt managers to delay bad-news disclosures and because of the nature of the earnings reporting process. Using returns to measure news, we find, consistent with our prediction, that earnings' informativeness relative to other sources is higher in bad-news quarters than in good-news quarters. Further, cross-sectional tests indicate that earnings' differential informativeness in bad-news quarters is more pronounced when managers do not issue short-horizon forecasts, information asymmetry is stronger, and managers are net sellers of stock. Our findings highlight that when factors such as litigation risk are not high enough to induce voluntary disclosures, earnings release the negative information that has not reached the market via alternative sources.
Scott Slotnik (Psychology)
Our research program aims to understand the nature of visual memory and visual attention using cognitive modeling (based on behavioral measures) and functional magnetic resonance imaging (fMRI, which involves modeling brain activity to identify regions associated with a specific cognitive process). We also developed and continually improve a script that is widely used in the field of Cognitive Neuroscience to correct for multiple comparisons in fMRI analysis. Optimization of this script requires conducting large sets of Monte Carlo simulations that systematically vary fMRI acquisition/analysis parameters such as volume dimensions, voxel size, spatial autocorrelation, individual voxel p-value, and familywise p-value.
Marc Snapper (Chemistry)
Our research focus is on introducing new chemical transforma¬tions, using these reactions to build complex molecules, and then using these compounds to study cellular function. To maintain competitiveness in the global chemical community, the introduction of new, more powerful chemical reactions continues to be an important endeavor in organic chemistry. Our efforts have been directed toward discovering better ways of constructing medium-ring-containing compounds. Using novel transformations that build molecular complexity rapidly have allowed for the efficient construction of seven- and eight-membered ring, containing natural products. Computational studies will be used to help predict the outcome of various synthetic approaches towards our target structures.
Philip E. Strahan (Finance)
Non financial firms have increased cash sharply following the financial crisis. Large firms and firms with access to bond markets have increased cash more than other firms, suggesting that financial constraints coupled with an increase in demand for liquidity helps explain these patterns.
Richard Sweeney (Economics)
This research project uses confidential microdata from the Energy Information Administration to study energy and environmental policy in US petroleum markets. One project looks at the impact of fuel content regulations stemming from the 1990 Clean Air Act Amendments on the US oil refining industry. A key feature of this industry is that economically interrelated markets experience differential regulation under the new rules, making estimation their impact complicated. We overcome this challenge by estimating a structural model of the industry and simulating policy counterfactuals for both regulated and unregulated markets simultaneously. A second project looks at the effects of the US crude oil export ban. The advent of fracking abruptly reversed a decades long decline in domestic oil production, yet physical and regulatory impediments have severely limited the transmission of this new crude. We estimate who benefited most from the fracking boom and compare the current regime to a world where these impediments are removed.
Mike Teodorescu (Information Systems)
Professor Teodorescu studies the behavior of firms pre and post the America Invents Act, particularly around strategic crafting of patent claims. In this project, acceleration of patent examination is another strategic choice taken at the firm level. This project employs natural language processing and machine learning tools.
Thanh Tran (Social Work)
The research team has on-going projects using large scale, nationally representative data to investigate wellness outcomes such as disability, self-rated health, depression, and other health and mental health indicators for Asian American populations. Key social determinants such as acculturation, lifestyle behaviors and other factors are examined across ethnic groups to determine how to best address health disparities and services in social work research and practice with this under-served population.
Jean-Baptiste Tristan (Computer Science)
We use machine learning to model the potential energy surface of molecules and optimize their geometry. In particular, we explore the design of Gaussian processes and neural networks that are invariant to actions of the orthogonal group, and try to scale them to the modeling of large molecules.
Chia-Kuang Tsung (Chemistry)
One strategy to mitigate anthropogenic climate change is to develop methods to transform carbon dioxide into value-added chemicals and fuels to help incentivize its capture. The hydrogenation of carbon dioxide to formic acid is an attractive route because of formic acid's potential use as both a fuel cell feedstock and as a C1 building block for more complex molecules; however requirements of high CO2 partial pressure and expensive noble metal catalysts prevent this reaction from reaching large scale implementation. A recently discovered pincer complex was found to have high activity for CO2 reduction at low pressure and using earth-abundant iron as its metal center. The catalyst is effective in an atmosphere of CO2 and nitrogen, but for real world implementation it must be resistant to oxygen, which presently degrades the catalyst. Encasing the complex in a porous metal organic framework (MOF) that allows the passage of CO2 but not of O2, such as ZIF-78 or CID-3, would provide the desired durability without impacting the catalysts inherent activity. Selection of an appropriate MOF material requires knowledge of not only spatially how the iron complex catalyst will fit within the pore cavities but also how it will interact chemically with functional groups in the MOF structure. A initio calculations with Gaussian software will allow us to screen many MOF candidates and identify the best material for making this multifunctional catalyst.
Robert Ulbricht (Economics)
This paper studies the dynamics of skill mismatch over the business cycle. We build a tractable directed search model, in which workers differ in skills along multiple dimensions and sort into jobs with heterogeneous skill requirements along those dimensions. Skill mismatch arises due to information and labor market frictions. Estimated to the U.S., the model replicates salient business cyclic properties of mismatch. We show that job transitions in and out of bottom job rungs, combined with career mobility of workers, are important to account for the empirical behavior of mismatch. The model suggests significant welfare costs associated with mismatch due to learning frictions.
Rosen Valchev (Economics)
This project seeks to understand the origins of nominal price stickiness, which is both a pervasive feature of the data, and a crucial ingredient in modern macroeconomic theoretical models Yet, standard models of price stickiness are at odds with certain robust empirical facts from micro price datasets. To address this, we explore a new, parsimonious theory of price rigidity, built around the idea of demand uncertainty, that is consistent with a number of salient micro facts. In the model, firms faces Knightian uncertainty about their competitive environment. They learn non-parametrically about the underlying, uncertain demand and make robust pricing decisions. The non-parametric learning leads to kinks in the expected profit function at previously observed prices, which generate price stickiness and a discrete price distribution. In addition, we show that when the ambiguity-averse firm worries that aggregate inflation is an ambiguous signal of the prices of its direct competitors in the short run, the rigidity becomes explicitly nominal in nature.
Tim van Opijnen (Biology)
Single Molecule, Real-Time (SMRT) Sequencing platforms (PacBio and Nanopore) have two major features that distinguish them from Second Generation Sequencing: 1) the production of very long reads (>10kb) making them ideal for genome assembly and closing of draft genomes; and 2) the ability to identify DNA modifications such as methylation. The goal of our project is to sequence the genomes of a representative number of strains of the pathogen Streptococcus pneumoniae.
Mary Walsh (Lynch School)
City Connects, situated in the Center for Optimized Student Support in the Lynch School, is an intervention that supports students to engage and learn in school by connecting each and every student in a school with the tailored set of prevention, intervention, and enrichment services s/he needs to thrive. Currently, the program is implemented in 100 schools across five states. Research has shown that City Connects’ approach to addressing out-of-school factors significantly improves academic performance. Ongoing evaluation that makes use of gold-standard analytic methods is a critical component of City Connects. Several current analyses rely on imputation and simulation methods with large data sets. For example, the organization is currently taking advantage of the opportunity to construct a natural experiment to evaluate the impact of City Connects on student academic achievement by relying on Monte Carlo simulations of an actual school assignment mechanism that took place across a public school system.
Dunwei Wang (Chemistry)
The Wang group's research is centered around artificial photosynthesis to mitigate problems caused by the usage of fossil fuels. We will use the cluster to understand the nature of charge transfer between catalyst and photoelectrodes. The understanding is an important piece of our effort to mimic photosynthesis in the lab, which will pave the way toward a green, sustainable solution to our energy needs.
Ziqiang Wang (Physics)
Our research group studies the fundamental physics of strongly correlated electronic materials with a special focus on that of the high temperature superconductors. Understanding the unconventional, complex, and emergent physical properties in these materials represents both the challenge and the vitality of condensed matter physics. The strong many-body correlations in these systems render the problem nonperturbative and the investigation of the possible electronic states of matter and the low energy excitations defies conventional perturbation approaches that use noninteracting electrons as a starting point. As a result, it is very difficult to study these materials by purely analytical means and numerical computations have played and continue to play a key role in this rapidly developing field. The computational component of the research projects in our group involves exact diagonalization, variational and quantum Monte Carlo simulations, and manipulations of large random matrices.
Eranthie Weerapana (Chemistry)
Our lab utilizes mass spectrometry to identify and quantify proteins from complex mixtures. We apply chemical probes to specifically enrich subsets of proteins based on activity or posttranslational modification-state for analysis by mass spectrometry using an LTQ-Orbitrap instrument. We are particularly interested in the functional significance of protein oxidation and glycosylation and seek to develop novel chemical proteomic technologies for quantitatively profiling these protein modifications. The BC research cluster will be used for data analysis software programs that correlate tandem mass spectra of peptides with amino acid sequences from protein databases. One such search algorithm, known as SEQUEST, cross correlates the observed tandem mass spectrum to theoretical spectra to identify the best candidate sequence match. Using a combination of chemistry, biology, mass spectrometry and bioinformatics, we hope to identify novel dysregulated protein activities implicated in a variety of patho physiological states.
Donglai Wei (Computer Science)
Our research focuses on developing computational algorithms for biomedical image analysis and natural video understanding. Current projects include brain map reconstruction, biomedical predictive modeling, and video instance segmentation.
Lai Wei (CSOM)
We consider a stochastic inventory model (under both backorder and lost-sales) with non-stationary demands, positive lead times, and sequential probabilistic service level constraints. This is a notoriously difficult problem to solve and, to date, not much progress has been made in understanding the structure of its optimal control, especially for the lost-sales inventory system. In this paper, we propose a simple order-up-to control, whose parameters can be calculated using the optimal solution of a deterministic approximation of the backorder inventory system, and show that it is asymptotically optimal for both the backorder and lost-sales systems in the regime of high service level requirement. This result contributes to the growing body of inventory literature that shows the near-optimality of simple heuristic controls. Moreover, it also gives credence to the use of deterministic approximation for solving complex inventory problems in practice, at least for applications where the targeted service level is sufficiently high. Our analysis for the lost-sales system involves the construction of an alternative backorder system whose expected total cost can be related to that of the analogous lost-sales system.
Yehua Wei (CSOM)
My research studies operational/decision problems under uncertainty, with a heavy emphasis on supply chain management. In my research, I use the Linux cluster to perform computational extensive tasks such as solving large-scale optimization problems and statistical analysis. My current research projects include analyzing the optimal supply chain risk mitigation strategies, designing efficient resource pooling networks, and computing (near) optimal policies for sequential decision problems.
Milena Wittwer (CSOM)
In traditional over-the-counter (OTC) markets, investors trade bilaterally through intermediaries referred to as dealers. An important regulatory question is whether to centralize OTC markets by shifting trades onto centralized platforms. We address this question in the context of the liquid Canadian government bond market. We document that dealers charge markups even in this market and show that there is a price gap between large investors who have access to a centralized platform and small investors who do not. We specify a model to quantify how much of this price gap is due to platform access and assess welfare effects. The model predicts that not all investors would use the platform even if platform access were universal. Nevertheless, the price gap would close by 32%--47%. Welfare would increase by 9%--30% because more trades are conducted by dealers who have high values to trade.
Zhijie Xiao (Economics)
Bootstrapping has attracted a lot of research attention in the last twenty years. It provides a convenient way of estimating the distribution of an estimator or test statistic by resampling the original data. In this project, we consider a prewhitened block bootstrap (PBB) method. The prewhitened block bootstrap combines the ideas of the parametric residual-based bootstrap and the nonparametric blockwise bootstrap. The stated idea of prewhitened block bootstrap is as follows: First, one prewhitens (prefilters) the original data to obtain a less dependent series; then block bootstrap is applied to the prewhitened (filtered) data, which has less dependence; finally the (block) bootstrapped data is recolored to produce a bootstrapped data set for the original series, and bootstrap estimation and inference procedures can be constructed based on the recolored data. Bootstrap is a very computational intensive method, high power computers are needed for this research project.
Nancy R. Xu (CSOM)
My current work focuses on the identification of the dynamics of risk aversion (price of risk) and economic uncertainties (amount of risk) and their effects on both domestic and international asset markets. My work requires complex estimation on the dynamics of economic and financial data. For example, in one of my projects, we study the joint dynamics of growth rates across 180 countries over the past 50 years by exploiting multivariate gamma shocks. In another work, I estimate and extract a global risk aversion measure by incorporation a large information set of financial and economic information around the world.
Xuan Ye (CSOM)
This paper investigates how the geographic organization of contributors influences open source software (OSS) production for business firms, a context in which contributors are incentivized exclusively by non-wage incentives. We test the hypothesis that the geography of OSS contributors plays an important role in explaining team productivity. We document that OSS contributors are more productive if they share time zones with project owners and collaborators. We identify this effect through an analysis of two million hourly contribution records over the last eight years on GitHub.com. Our preliminary findings suggest that a local, “home” bias may exist in open source production, and the success of firms’ open source projects may be constrained by a localized pool of talent, muchlike with offline production.
Hanyi Yi (CSOM)
This project studies whether and how shocks to the supply of municipal credit affect both the quantity and quality of local public goods provision, and through this channel, resident migration. I use a difference-in-differences approach based on a policy that unintendedly decreased bank investment in municipal bonds, and an amendment to the policy that partially reversed its effects. I assemble a massive dataset that consists of government-level data on spending and investment, county-level data on the quality of public goods, and individual-level data on geographic mobility.
Liane Young (Psychology)
Our research group studies the cognitive and neural basis of human moral judgment. Our current research focuses on the role of theory of mind, mind attribution, and emotions in moral judgment and behavior, as well as individual and cultural differences in moral cognition. We employ methods of social psychology and cognitive neuroscience, including functional magnetic resonance imaging (fMRI).
Ilija Zeljkovic (Physics)
Modeling of Coulomb blockade effects in tunneling measurements
Peter Zhang (Chemistry)
Research in the Zhang group focuses on the development of metalloradical catalysis (MRC) as a new concept to develop general approaches for controlling reactivity and selectivity of radical reactions. The current projects involve the design of Co(II)-based metalloradical catalysts for enantioselective olefin aziridination/cyclopropanation and C-H alkylation/amination reactions. The Linux Cluster is used for performing molecular modeling of the catalytic radical processes. Theoretical investigations carried out with Gaussian and/or Jaguar will be invaluable to the design and optimization of these catalytic radical processes.
Huiqing (Jane) Zhou [Chemistry]
The Zhou Lab is interested in understanding the recognition mechanism between RNA chemical modification reader protein and the modified RNA, and in engineering designer reader proteins to recognize chemically modified RNA. The cluster will be used to predict and assess binding energetics between modified RNA and protein residues.
The andromeda cluster was installed in the Fall of 2020 and upgraded in January 2022. In addition to 500 TB of storage, and an interactive node, it has 176 compute nodes with a total of 9792 cores and 7 GPU nodes.
- 92 compute nodes. Each has 2 24-core Intel Xeon Platinum 8260 CPU (2.40GHz) sharing 192 GB of memory.
- 84 compute nodes. Each has 2 32-core Intel Xeon Platinum 8352Y CPU (2.20GHz) sharing 256 GB of memory.
- 3 GPU nodes. Each with 2 24-core Intel Xeon Platinum 8260 CPU (2.40GHz) sharing 192 GB and 4 Nvidia V100 GPUs.
- 4 GPU nodes. Each with 2 32-core Intel Xeon Platinum 8352Y CPU (2.20GHz) sharing 256 GB and 4 Nvidia A100 GPUs.
The sirius cluster was installed in the Summer of 2017 and additional nodes were added in the Summer of 2018 and in January 2019. In addition to 600 TB of storage, and an interactive node, it has 134 nodes with a total of 4,488 cores.
- 66 compute nodes. Each nodes has two 14-core Intel Xeon E5-2680 v4 processors (2.40GHz) sharing 128 GB of memory.
- 66 compute nodes. Each node has two 20-core Intel Xeon Gold 6148 processors (2.40GHz) sharing 192 GB of memory.
- 2 nodes with two 20-core Intel Xeon Gold 6148 processors (2.40GHz) sharing 192 GB of memory and 4 Nvida V100 GPUs.
Request a New Linux Cluster Account
Requests for new Linux accounts must come from a faculty member. We will try for an 8 business-hour turnaround time (after all information has been received), and try to accommodate special requests when accounts are needed sooner. To get a new account, send email to email@example.com with the following information:
- The name of the faculty member sponsoring the account.
- If this is the first account for the faculty member, or if it is a new group, please include a short abstract describing the work on the cluster. This abstract will be displayed on the Research Projects website.
- The name of account holder, and their BC username, if applicable.
Once an abstract has been submitted, a faculty member may request accounts for members of their research group. Members of the research group are not restricted to the Boston College community. We will create accounts for collaborators at institutions other than Boston College.
Once a year, we will ask each research group to update their abstract and update, if necessary, the accounts for the group. New groups and new accounts may be created at any time.
We also ask each research group to provide us:
- A list of the publications in which the cluster supported their work.
- A list of grants submitted in which the cluster was mentioned.
- A list of funded grants in which the cluster was mentioned.
- We need the duration of the grant, the amount of the grant, and the funding agency when we request information about grant activity that involves the cluster. Please send this information as it becomes available to firstname.lastname@example.org.
We expect all account holders to abide by the Cluster Policies:
We expect users to follow common policies for the use of shared computers, including:
- No shared accounts. The standard UNIX group structure can easily accommodate sharing of files within a group. If you have special needs, please contact us. We should be able to accomodate these needs.
- Boston College's Computing Policies and Guidelines proper use policies apply to accounts on the cluster.