Project: Designing next generation batteries with AI

Description: A sustainable energy economy requires an immediate shift to renewable
technologies, and batteries are vital to counterbalance the intermittency of solar and wind.
Batteries consist of an anode, cathode, and an electrolyte; while the anode/cathode selection
determines the theoretical energy density, the electrolyte choice often controls battery lifetime.
Furthermore, electrolyte selection can determine how fast a battery charges, battery operation
temperature, and safety. The conflicting requirements for electrolytes often mean that
electrolytes are discovered using trial-and-error approaches; slowing down research progress and
time to commercialization. Here, data science and the ability of AI to extract meaning from
complex data is of vital interest. We have a multitude of data science + battery projects. One
example project involves the development of natural language processing (NLP) and computer
vision algorithms to extract battery data from the literature to build machine learning models.
Although the scientific literature contains decades of relevant battery information, utilization of
this data has been stifled by the lack of NLP tools capable of digesting the information. The
second example project is focused on developing ML tools and applying these tools to discover
novel electrolyte compounds for novel battery chemistries. Combining your data science
expertise with the battery (and materials) domain knowledge that you will obtain from our
laboratory will provide you with a skillset to exploit the battery sector that is expected to grow to
over 100 billion dollars by 2025.

Mentor: Chibueze Amanchukwu is a Neubauer Family Assistant Professor in the Pritzker School of
Molecular Engineering at the University of Chicago, and holds a joint appointment in the
Chemical Sciences and Engineering Division at Argonne National Laboratory. His research
involves the design, synthesis, and understanding of ion transport in electrolytes for batteries and
electrocatalytic applications. His team is especially interested in correlating bulk solvation
properties to electrochemical interfacial phenomena, and they use data science tools coupled
with synthesis and novel characterization approaches to answer these questions. He obtained his
B.S. in chemical engineering at Texas A&M University, PhD in chemical engineering as a
NDSEG Fellow at MIT and was a TomKat Center Postdoctoral Fellow at Stanford University.

Projects: Data-driven Environmental Enforcement

The Energy and Environment Lab invites a postdoc to collaborate on a suite of projects that leverage advances in monitoring technology and machine learning approaches to inform environmental policy, under the mentorship of Michael Greenstone, the Milton Friedman Distinguished Service Professor in Economics, the College, and the Harris School; Director of the Energy and Environment Lab, the Becker Friedman Institute, and the Energy Policy Institute at Chicago.

Congestion and Traffic Safety

Many cities across the United States have adopted Vision Zero, the policy goal of eliminating all traffic-related deaths and serious injuries. But what are the costs of achieving Vision Zero and what are the most efficient policy instruments to get there? Monitoring technologies offer the potential to revolutionize urban policy by providing governments with big data to inform policymaking. As part of NYC’s Vision Zero, the Department of Transportation is more than quintupling the number of speed cameras in the city. Leveraging our access to unique data of taxi, for-hire vehicle, and city fleet trips to model the impacts of new traffic cameras on vehicle crashes, slowdowns, and congestion spillovers. The post-doc would utilize large administrative datasets on camera enforcement, vehicle crashes, segment-level traffic speeds, and high-resolution driver behavior, to help measure the costs and benefits of enforcement strategies for fatality/injury reduction to inform optimal policy for urban traffic safety.

Leveraging Satellite Data to Reduce Oil & Gas Methane Emissions

The meteoritic rise of shale oil and gas (O&G) drilling in the United States poses significant challenges for reducing greenhouse gas emissions. The methane emitted has around 30 times greater short-term global warming potential than CO2, contributing aggressively to climate change. Reliable estimates of emitted methane are essential to fully understand and mitigate the environmental threat presented by shale drilling. While some estimates suggest that approximately 2.3% of gross natural gas production is leaked per year, accurate monitoring of emissions remains extremely challenging. Currently, regulators visit individual facilities to measure emissions; but due to budgetary constraints and a fast-growing industry inspector can visit only a fraction of the facilities each year. This project will leverage a wealth of administrative data and novel remote sensing data from recently-launched satellites to estimate facility-level methane emissions. Leveraging these unique data and state-of-the-art machine learning techniques, the project will help regulators re-design their monitoring and enforcement strategy to realize improvements in regulatory efficiency and reductions in greenhouse gases.

Beyond Inspection Targeting: Deterrence through Machine Learning

Building on a three-year partnership with the Environmental Protection Agency (EPA), this project aims to scale a machine learning-driven framework across inspection targeting programs at EPA. The Clean Water Act (CWA) is one program where data-driven inspection targeting can directly influence environmental policy. Using state-of-the-art machine learning models, we can generate risk scores for the likelihood individual firms will violate CWA standards, and use these model-generated risk scores to study facility compliance behavior and identify the most effective approaches to deterrence in a randomized field trial.

Mentor: Michael Greenstone, Milton Friedman Distinguished Service Professor in Economics, the College, and the Harris School, University of Chicago; Director, Becker Friedman Institute for Research in Economics; Director, Energy Policy Institute at the University of Chicago (EPIC); Director, Tata Center for Development at the University of Chicago

Language learning has come to be a central theme in both cognitive science and artificial intelligence. The nature of language learning has long been a topic of interest for cognitive scientists, and machine learning has begun to dominate natural language processing (NLP) in modern AI. NLP systems have benefited tremendously from machine learning. However, the learning systems developed using these procedures often don’t achieve the efficiency and robustness of human language acquisition. Insights from language acquisition have the potential to help address this problem. But there are two critical challenges in exploring this possibility: (1) identifying the innate learning biases that enable fast, robust language learning in humans, and (2) determining how to translate theoretical insights about these biases into effective implementation for learning in NLP systems. This project will tackle both of these issues, making use of insights from special linguistic populations.

Our first challenge –– identifying innate language learning predispositions –– is driven by the fact that most children are exposed to linguistic input from birth, making it difficult to disentangle innate characteristics versus characteristics that are rapidly learned from input. The rare cases in which children do not have usable linguistic input can help here by allowing us make important headway in identifying these predispositions. Congenitally deaf children who cannot learn the spoken language that surrounds them, and who have not been exposed to sign language by their hearing families, are in the unique situation of being without language input early in life. These children use their hands to communicate –– they gesture –– and those gestures (called “homesigns”) take on many, but not all, of the forms and functions of languages that have been handed down from generation to generation. The properties of these naturally-arising gestures provide evidence for the nature of linguistic predispositions independent of input.

Drawing candidate biases from homesign, we will then tackle the second challenge –– incorporating biases into machine learning systems –– by systematic testing of models against real-world child language acquisition data. The goal of this phase will be to identify effective means of instantiating proposed human biases, and to test whether models incorporating these biases will successfully simulate the learning trajectories exhibited by children. Models with the proposed biases will be compared against minimally-different baseline models lacking the biases; stronger fit to human data will be taken as support that the biases are actual human predispositions. An important priority of this phase will be to balance scientific and engineering needs –– to maintain transparency of the models’ cognitive implications and to simulate human learning patterns as closely as possible, but also to use models that will interface smoothly with modern NLP systems, with promise to scale to larger datasets and broader domains.

Mentor: Allyson Ettinger, Assistant Professor, Department of Linguistics

Dr. Allyson Ettinger’s research is focused on language processing in humans and in artificial intelligence systems, motivated by a combination of scientific and engineering goals. For studying humans, her research uses computational methods to model and test hypotheses about mechanisms underlying the brain’s processing of language in real time. In the engineering domain, her research uses insights and methods from cognitive science, linguistics, and neuroscience in order to analyze, evaluate, and improve natural language understanding capacities in artificial intelligence systems. In both of these threads of research, the primary focus is on the processing and representation of linguistic meaning.

Project: Operational Analytics for Communications Systems

Many applications of machine learning to computer communications systems such as the Internet rely on models that are trained offline, on snapshots of data. Yet, in many operational systems, data arrives as a continuous stream—often as a timeseries, and decisions must be made on short timescales (e.g., milliseconds). In operational systems, designers must face difficult challenges and design tradeoffs concerning encoding of timeseries data, efficiently labeling large quantities of data, distinguishing anomalous activity from model drift, and trading off model accuracy versus model or feature complexity. In this research project, we will explore these challenges in the context of networked systems. Possible avenues include device identification and anomaly detection in industrial control systems and consumer “IoT” smart homes; streaming video quality estimation; content moderation on social media platforms (e.g., Facebook); and security vulnerability detection in networked systems.

Mentor: Nick Feamster, Neubauer Professor of Computer Science; Director, Center for Data and Computing

Nick Feamster is Neubauer Professor of Computer Science and the Director of Center for Data and Computing (CDAC) at the University of Chicago. Previously, he was a full professor in the Computer Science Department at Princeton University, where he directed the Center for Information Technology Policy (CITP); prior to Princeton, he was a full professor in the School of Computer Science at Georgia Tech.

His research focuses on many aspects of computer networking and networked systems, with a focus on network operations, network security, and censorship-resistant communication systems. He received his Ph.D. in Computer science from MIT in 2005, and his S.B. and M.Eng. degrees in Electrical Engineering and Computer Science from MIT in 2000 and 2001, respectively. He was an early-stage employee at Looksmart (acquired by AltaVista), where he wrote the company’s first web crawler; and at Damballa, where he helped design the company’s first botnet-detection algorithm.

Nick is an ACM Fellow. He received the Presidential Early Career Award for Scientists and Engineers (PECASE) for his contributions to cybersecurity, notably spam filtering. His other honors include the Technology Review 35 “Top Young Innovators Under 35” award, the ACM SIGCOMM Rising Star Award, a Sloan Research Fellowship, the NSF CAREER award, the IBM Faculty Fellowship, the IRTF Applied Networking Research Prize, and award papers at ACM SIGCOMM (network-level behavior of spammers), the SIGCOMM Internet Measurement Conference (measuring Web performance bottlenecks), and award papers at USENIX Security (circumventing web censorship using Infranet, web cookie analysis) and USENIX Networked Systems Design and Implementation (fault detection in router configuration, software-defined networking). His seminal work on the Routing Control Platform won the USENIX Test of Time Award for its influence on Software Defined Networking.

Project: Data Markets and the Economics of Data

Data has been called the new oil, and it is certainly true that the sheer volume, variety, and velocity of data being produced and stored is generating enormous value for the individuals and organizations that know how to tap into and refine it. But data is not just another commodity inhabiting an economic and social system; it has given rise to an entirely new economy and society. This research project will investigate theoretical, empirical, and technological foundations of the new data and artificial intelligence economy comprising emerging data markets, data integration and the transformational services that fund these changes. Sub-projects include research into: pricing models for data, architecture of data sharing platforms, data discovery, and data provenance and integration.

Mentor Bio: 

Michael J. Franklin is the inaugural holder of the Liew Family Chair of Computer Science. An authority on databases, data analytics, data management and distributed systems, he also serves as senior advisor to the provost on computation and data science.

Previously, Franklin was the Thomas M. Siebel Professor of Computer Science and chair of the Computer Science Division of the Department of Electrical Engineering and Computer Sciences at the University of California, Berkeley. There, he co-founded Berkeley’s Algorithms, Machines and People Laboratory (AMPLab), a leading academic big data analytics research center. The AMPLab won a National Science Foundation CISE “Expeditions in Computing” award, which was announced as part of the White House Big Data Research initiative in March 2012, and received support from over 30 industrial sponsors. AMPLab created industry-changing open source Big Data software including Apache Spark and BDAS, the Berkeley Data Analytics Stack. At Berkeley, he also served as an executive committee member for the Berkeley Institute for Data Science, a campus-wide initiative to advance data science environments.

An energetic entrepreneur in addition to his academic work, Franklin founded and became chief technology officer of Truviso, a data analytics company acquired by Cisco Systems. He serves on the technical advisory boards of various data-driven technology companies and organizations.

Franklin is a Fellow of the Association for Computing Machinery and a two-time recipient of the ACM SIGMOD (Special Interest Group on Management of Data) “Test of Time” award. His many other honors include the outstanding advisor award from Berkeley’s Computer Science Graduate Student Association. He received the Ph.D. in Computer Science from the University of Wisconsin in 1993, a Master of Software Engineering from the Wang Institute of Graduate Studies in 1986, and the B.S. in Computer and Information Science from the University of Massachusetts in 1983.

Homepage

Language learning has come to be a central theme in both cognitive science and artificial intelligence. The nature of language learning has long been a topic of interest for cognitive scientists, and machine learning has begun to dominate natural language processing (NLP) in modern AI. NLP systems have benefited tremendously from machine learning. However, the learning systems developed using these procedures often don’t achieve the efficiency and robustness of human language acquisition. Insights from language acquisition have the potential to help address this problem. But there are two critical challenges in exploring this possibility: (1) identifying the innate learning biases that enable fast, robust language learning in humans, and (2) determining how to translate theoretical insights about these biases into effective implementation for learning in NLP systems. This project will tackle both of these issues, making use of insights from special linguistic populations.

Our first challenge –– identifying innate language learning predispositions –– is driven by the fact that most children are exposed to linguistic input from birth, making it difficult to disentangle innate characteristics versus characteristics that are rapidly learned from input. The rare cases in which children do not have usable linguistic input can help here by allowing us make important headway in identifying these predispositions. Congenitally deaf children who cannot learn the spoken language that surrounds them, and who have not been exposed to sign language by their hearing families, are in the unique situation of being without language input early in life. These children use their hands to communicate –– they gesture –– and those gestures (called “homesigns”) take on many, but not all, of the forms and functions of languages that have been handed down from generation to generation. The properties of these naturally-arising gestures provide evidence for the nature of linguistic predispositions independent of input.

Drawing candidate biases from homesign, we will then tackle the second challenge –– incorporating biases into machine learning systems –– by systematic testing of models against real-world child language acquisition data. The goal of this phase will be to identify effective means of instantiating proposed human biases, and to test whether models incorporating these biases will successfully simulate the learning trajectories exhibited by children. Models with the proposed biases will be compared against minimally-different baseline models lacking the biases; stronger fit to human data will be taken as support that the biases are actual human predispositions. An important priority of this phase will be to balance scientific and engineering needs –– to maintain transparency of the models’ cognitive implications and to simulate human learning patterns as closely as possible, but also to use models that will interface smoothly with modern NLP systems, with promise to scale to larger datasets and broader domains.

Mentor: Susan Goldin-Meadow, Beardsley Ruml Distinguished Service Professor in the Department of Psychology and Committee on Human Development

Susan Goldin-Meadow is the Beardsley Ruml Distinguished Service Professor in the Department of Psychology and Committee on Human Development at the University of Chicago. A year spent at the Piagetian Institute in Geneva while an undergraduate at Smith College piqued her interest in the relationship between language and thought, interests she continued to pursue in her doctoral work at the University of Pennsylvania (Ph.D. 1975). At Penn and in collaboration with Lila Gleitman and Heidi Feldman, she began her studies exploring whether children who lack a (usable) model for language can nevertheless create a language with their hands. She has found that deaf children whose profound hearing losses prevent them from learning the speech than surrounds them, and whose hearing parents have not exposed them to sign, invent gesture systems which are structured in language-like ways. This interest in how the manual modality can serve the needs of communication and thinking led to her current work on the gestures that accompany speech in hearing individuals. She has found that gesture can convey substantive information – information that is often not expressed in the speech it accompanies. Gesture can thus reveal secrets of the mind to those who pay attention.

Professor Goldin-Meadow’s research has been funded by the National Science Foundation, the Spencer Foundation, the March of Dimes, the National Institute of Child Health and Human Development, and the National Institute of Neurological and Communicative Disorders and Stroke. She has served as a member of the language review panel for NIH, has been a Member-at-Large to the Section on Linguistics and Language Science in AAAS, and was part of the Committee on Integrating the Science of Early Childhood Development sponsored by the National Research Council and the Institute of Medicine and leading to the book Neurons to Neighborhoods. She is a Fellow of AAAS, APS, and APA (Divisions 3 and 7). In 2001, she was awarded a Guggenheim Fellowship and a James McKeen Cattell Fellowship which led to her two recently published books, Resilience of Language and Hearing Gesture. In addition, she edited Language in Mind: Advances in the Study of Language and Thought in collaboration with Dedre Gentner. She has received the Burlington Northern Faculty Achievement Award for Graduate Teaching and the Llewellyn John and Harriet Manchester Quantrell Award for Excellence in Undergraduate Teaching at the University of Chicago. She is currently the President of the Cognitive Development Society and the editor of the new journal sponsored by the Society for Language Development, Language Learning and Development. Professor Goldin-Meadow also serves as chair of the developmental area program.

Chenhao Tan is an assistant professor at the Department of Computer Science and the UChicago Data Science Institute. His main research interests include language and social dynamics, human-centered machine learning, and multi-community engagement. He is also broadly interested in computational social science, natural language processing, and artificial intelligence.

Homepage.

Project: Machine Learning for Physical Systems
Much of machine learning is focused on recognizing patterns and making predictions based on training data. However, in many physical science settings, the complexity of the task is too high for effective learning giving the amount of available data. In these settings, it is essential to incorporate knowledge of the underlying physical system to mitigate the effect of limited data. Examples include using a combination of training data and models of a CT scanners operation to develop better medical image reconstruction methods, leveraging both observational and simulated data to develop better climate predictions, and building deep learning-based surrogate models for computationally demanding PDE-based simulators of physical systems. While there are isolated examples of successes in these regimes, little is known on a fundamental level. What are optimal machine learning methods that leverage both training data and physical models? How does sample complexity scale with the type of physical system and the accuracy of our models? Which kinds of PDE models are most amenable to deep surrogate models? This project will focus on developing new methodology and theory for machine learning for physical system that will address these and other open problems.

Mentor: Rebecca Willett, Professor, Statistics, Computer Science, and the College

Rebecca Willett is a Professor of Statistics and Computer Science at the University of Chicago. Her research is focused on machine learning, signal processing, and large-scale data science. She completed her PhD in Electrical and Computer Engineering at Rice University in 2005 and was an Assistant then tenured Associate Professor of Electrical and Computer Engineering at Duke University from 2005 to 2013. She was an Associate Professor of Electrical and Computer Engineering, Harvey D. Spangler Faculty Scholar, and Fellow of the Wisconsin Institutes for Discovery at the University of Wisconsin-Madison from 2013 to 2018. Willett received the National Science Foundation CAREER Award in 2007, is a member of the DARPA Computer Science Study Group, and received an Air Force Office of Scientific Research Young Investigator Program award in 2010. Willett has also held visiting researcher or faculty positions at the University of Nice in 2015, the Institute for Pure and Applied Mathematics at UCLA in 2004, the University of Wisconsin-Madison 2003-2005, the French National Institute for Research in Computer Science and Control (INRIA) in 2003, and the Applied Science Research and Development Laboratory at GE Healthcare in 2002.

Project: Causality for Credible AI. The backbone of modern ML is evaluation on randomly held out data. However, models with strong held out performance frequently still exhibit disturbing failures. For example, they can degrade when deployed out of domain, learn to depend on apparently irrelevant input features, or change substantively in response to tiny variations in training procedure (e.g., changing the random seed). The aim of this project is to develop methods to mitigate this kind of failure by baking in causal knowledge to model design and evaluation. For example, we may know changing certain input features shouldn’t change predictions (e.g., your medical diagnosis shouldn’t change if we edit your zipcode), or we may know something about common causal structure between domains (rain should affect a self-driving car the same in Chicago and New York). The challenge here is to translate domain knowledge and desiderata into formal requirements, and then determine how to enforce and measure these using available data.

Bio: I am an assistant professor of Statistics and Data Science at the University of Chicago and a research scientist at Google Cambridge. My recent work revolves around the intersection of machine learning and causal inference, as well as the design and evaluation of safe and credible AI systems. Other noteable areas of interests include network data, and the foundations of learning and statistical inference.

I was previously a Distinguished Postdoctoral Researcher in the department of statistics at Columbia University, where I worked with the groups of David Blei and Peter Orbanz. I completed my Ph.D. in statistics at the University of Toronto, where I was advised by Daniel Roy. In a previous life, I worked on quantum computing at the University of Waterloo. I won a number of awards, including the Pierre Robillard award for best statistics thesis in Canada.

Project: Creating Personalized Incentives to Drive Diabetes Patients’ Behavior

Physical exercise can product a significant health benefit for diabetics, but not all patients have the same natural inclinations for exercise. We are working to develop a heterogenous treatment model that would create individualized incentives to help diabetes patients succeed at an exercise regimen. Come have an outsized impact on this project in its early stages as we formulate initial data needs and begin sourcing additional measures with which to create our model.

Mentor: Rebecca Dizon-Ross, Associate Professor of Economics and Charles E. Merrill Faculty Scholar, Booth School of Business

Rebecca Dizon-Ross is a development economist with an interest in human capital. Much of her current work is on the demand-side, aiming to understand the determinants of households’ investments in health and education.

Before joining Booth, Dizon-Ross was a Prize Fellow in Economics, History, and Politics at Harvard University and a Postdoctoral Fellow in the Abdul Latif Jameel Poverty Action Lab at the Massachusetts Institute of Technology. She received a Ph.D. in Economics from Stanford University and a B.A. (summa cum laude) from Harvard University. Prior to graduate school, she worked as an analyst at McKinsey & Co.

Project: Nightingale Project

Machine learning, we are told, will transform medical diagnosis and patient care: by integrating ‘big data’ on patients’ history and physiology, algorithms can dramatically improve the quality of doctors’ decisions, with the potential both to reduce waste, avoid misdiagnosis, and produce breakthrough discoveries. For example, if massive datasets of ECG waveforms could be linked to national mortality registries, we could supercharge the current research, and find better, more consistent ways to allocate life-saving defibrillators. But most clinical data like this is siloed by different institutions and unavailable to researchers. Further complicating things, in order to protect patient privacy, public medical datasets are almost universally limited to a single, easily de-identified stream of information, like a set of X-rays.

The goal of the Nightingale Project in the Booth Center for Applied AI is to gather and share just the sort of rich, multidimensional data needed to feed AI-enabled discovery. Work with a team of engineers, data analysts, and medical experts to build a secure platform that can warehouse curated de-identified clinical datasets linked to ground truth outcomes. Using this initiative as a proof of concept to develop other privacy tools, own and work through the de-identification, sharing, and privacy components of large-scale datasets — work that will include writing models and creating security-related challenges — with the goal of making the data available to researchers securely.

Mentor: Sendhil Mullainathan, Faculty Director, Center for Applied AI, Roman Family University Professor of Computation and Behavioral Science, Chicago Booth

Sendhil Mullainathan is the Roman Family University Professor of Computation and Behavioral Science at Chicago Booth. His current research uses machine learning to understand complex problems in human behavior, social policy, and especially medicine, where computational techniques have the potential to uncover biomedical insights from large-scale health data. He currently teaches a course on Artificial Intelligence.

In past work he has combined insights from economics and behavioral science with causal inference tools—lab, field, and natural experiments—to study social problems such as discrimination and poverty. Papers include: the impact of poverty on mental bandwidth; how algorithms can improve on judicial decision-making; whether CEO pay is excessive; using fictitious resumes to measure discrimination; showing that higher cigarette taxes makes smokers happier; and modeling how competition affects media bias.

Mullainathan enjoys writing. He recently co-authored Scarcity: Why Having too Little Means so Much and writes regularly for the New York Times. Additionally, his research has appeared in a variety of publications including the Quarterly Journal of Economics, Science, American Economic Review, Psychological Science, the British Medical Journal, and Management Science.

Mullainathan helped co-found a non-profit to apply behavioral science (ideas42), co-founded a center to promote the use of randomized control trials in development (the Abdul Latif Jameel Poverty Action Lab), serves on the board of the MacArthur Foundation, has worked in government in various roles, is affiliated with the NBER and BREAD, and is a member of the American Academy of Arts and Sciences.

Prior to joining Booth, Mullainathan was the Robert C. Waggoner Professor of Economics in the Faculty of Arts and Sciences at Harvard University, where he taught courses about machine learning and big data. He began his academic career at the Massachusetts Institute of Technology.

Mullainathan is a recipient of the MacArthur “Genius Grant,” has been designated a “Young Global Leader” by the World Economic Forum, was labeled a “Top 100 Thinker” by Foreign Policy Magazine, and was named to the “Smart List: 50 people who will change the world” by Wired Magazine (UK).

Project: Early Childhood Metric Initiative

Early childhood suffers from a lack of quantifiable data that is easy to collect at scale. Work with a team of engineers, computer scientists, and early childhood experts to build a non-intrusive, wearable technology that leverages machine-learning to collect real-time, real-world data to measure young children’s early language environments (i.e. the quantity and quality of language interactions they are exposed to). As part of the project, develop machine learning algorithms and models to analyze the large-scale adult-child interaction data collected from this wearable technology. This large dataset will enable researchers and practitioners to better understand the relationships between family demographics, parental inputs, and child outcomes and identify effective approaches, as well as enable policymakers to hone in on the most effectives programs and policies to enhance children’s early language environments. This audio dataset will also allow experts in natural language processing to develop and refine speech processing algorithms.

Mentor: Dana Suskind, MD, Professor of Surgery and Pediatrics, Director, Pediatric Cochlear Implantation Program, Co-Director, Thirty Million Words (TMW) Center for Early Learning + Public Health

Dana Suskind, MD, is a pediatric otolaryngologist who specializes in hearing loss and cochlea implantation. She directs the University of Chicago Medicine’s Pediatric Hearing Loss and Cochlear Implant program.

Recognized as a national thought leader in early language development, Dr. Suskind has dedicated her research and clinical life to optimizing foundational brain development and preventing early cognitive disparities and their lifelong impact. She is founder and co-director of the TMW Center for Early Learning + Public Health, which aims to create a population-level shift in the knowledge and behavior of parents and caregivers to optimize the foundational brain development in children from birth to five years of age, particularly those born into poverty.

Her book “Thirty Million Words: Building a Child’s Brain” was published in 2015.

Dr. Suskind has received several awards for her work, including the Weizmann Women for Science Vision and Impact award, the SENTAC Gray Humanitarian Award, the LENA Research Foundation Making a Difference Award, the Chairman’s Award from the Alexander Graham Bell Association for the Deaf and Hard of Hearing in 2018, and the John D. Arnold, MD Mentor Award for Sustained Excellence from the Pritzker School of Medicine.

Project: Pediatric Cancer Data Commons

Collecting, aggregating, harmonizing, and sharing data from children with cancer is essential to making new discoveries and developing new cures. Too often, data are siloed and disconnected, drastically reducing the usefulness of these valuable resources. The Pediatric Cancer Data Commons (PCDC) at UChicago brings together researchers from around the world with the goal of building data dictionaries for all types of pediatric cancer. Consensus data models are balloted with experts from around the world, including clinicians, ontologists/taxonomists, statisticians, and data scientists. The resulting dictionary is used for harmonizing data from completed clinical trials and is subsequently leveraged as a framework for collecting data on new studies. The data are made available to the worldwide research community through a public-facing cohort discovery tool. Data are further connected to other sources through common identifiers, allowing novel new data sets to be developed for research and discovery.

Potential areas of research include: ontology development and data dictionary creation, data harmonization, automated methods of metadata extraction and data ingestion, development and deployment of novel data visualization tools and analytics, data governance and provenance methods and tools, developing novel methods of combining disparate data sets, and developing analytic methods for new modes of risk stratification. Experience with clinical data is preferred but not required.

Mentor: Samuel L. Volchenboum, Associate Professor of Pediatrics & Associate Chief Research Informatics Officer, UChicago Medicine

Samuel L. Volchenboum, MD, PhD, MS, is an expert in pediatric cancers and blood disorders. He has a special interest in treating children with neuroblastoma, a tumor of the sympathetic nervous system.

In addition to caring for patients, Dr. Volchenboum studies ways to harness computers to enable research and foster innovation using large data sets. He directs the development of the International Neuroblastoma Risk Group Database project, which connects international patient data with external information such as genomic data and tissue availability. The Center he runs provides computational support for the Biological Sciences Division at the University of Chicago, including high-performance computing, applications development, bioinformatics, and access to the clinical research data warehouse.

Project: Measuring Messages About Race and Gender

Early influences that depress children’s beliefs about their own ability can lead to lower educational achievement and persistent disadvantage. In particular, receiving negative messages about gender- and race-specific levels of ability have played a role in generating disadvantage for women and minorities. Children are particularly vulnerable to negative messages about race and gender, as their beliefs about their own capacities are highly malleable.

This project aims to improve how we estimate the extent and implications of children’s exposure to race- and gender-coded messages. The initial goal of the project will be to develop, verify, apply, and disseminate new methods of human-directed, machine-implemented content analysis focused on measuring implicit messages about race and gender in visual content. We anticipate that the tools we develop will catalyze new research in fields such as computational science and social science, and will advance our understanding of the extent of these messages and their contribution to inequality.

Mentors: Anjali Adukia and Hakizumwami Birali Runesha

Anjali Adukia is an assistant professor at the University of Chicago Harris School of Public Policy and the College. In her work, she is interested in understanding how to reduce inequalities such that children from historically disadvantaged backgrounds have equal opportunities to fully develop their potential.  Her research is focused on understanding factors that motivate and shape behavior, preferences, attitudes, and educational decision-making, with a particular focus on early-life influences.  She examines how the provision of basic needs—such as safety, health, justice, and representation—can increase school participation and improve child outcomes in developing contexts.

Adukia completed her doctoral degree at the Harvard University Graduate School of Education, with an academic focus on the economics of education. Her work has been funded from organizations such as the William T. Grant Foundation, the National Academy of Education, and the Spencer Foundation.  Her dissertation won awards from the Association for Public Policy Analysis and Management (APPAM), Association for Education Finance and Policy (AEFP), and the Comparative and International Education Society (CIES). Adukia received recognition for her teaching from the University of Chicago Feminist Forum.  She completed her masters of education degrees in international education policy and higher education (administration, planning, and social policy) from Harvard University and her bachelor of science degree in molecular and integrative physiology from the University of Illinois at Urbana-Champaign.  She is a faculty research fellow of the National Bureau of Economic Research and a faculty affiliate of the University of Chicago Education Lab and Crime Lab.  She is on the editorial board of Education Finance and Policy.  She was formerly a board member of the Young Nonprofit Professionals Network – San Francisco Bay Area. She continues to work with non-governmental organizations internationally, such as UNICEF and Manav Sadhna in Gujarat, India.

Hakizumwami Birali Runesha is the Director of Research Computing for the University of Chicago, where he provides leadership and vision for advancing all aspects of research computing strategies at the University. He is responsible for the design, configuration, and administration of centrally managed High-Performance Computing (HPC) systems and related services across the University. In addition, he provides access to advanced technical expertise, user support, advice and training, and access to the University’s HPC facility to the research community.

Runesha is a seasoned professional who brings to the University of Chicago HPC management leadership and more than 17 years of experience in high performance computing and scientific software development. He earned his M.S. and Ph.D. in Civil engineering at Old Dominion University. Prior to joining the University of Chicago, he served as Director of Scientific Computing and Applications at the University of Minnesota Supercomputing Institute (MSI) managing the scientific computing, biological computing, visualization and application development groups. In addition to overseeing strategic planning of HPC resources and leading annual procurement of supercomputing resources at MSI, Runesha created the MSI Application software development group and the MSI Scientific Data Management Laboratory to meet the evolving data management and database development needs of university researchers. Prior to joining the University of Minnesota, he was a research scholar at the Hong Kong University of Science and Technology developing parallel computing algorithms for engineering applications, a research associate for the Multidisciplinary Parallel-Vector Computer Center at Old Dominion University and an Assistant Professor at the University of Kinshasa.

Runesha has developed open source software programs and fast parallel solvers for large-scale finite element applications. He served as principal investigator on a number of research grants and is the author of a number of journal articles, proceedings and conference papers. He has given many invited talks, seminars, courses, and workshops on various HPC topics.

Project: Corporate influence on the rulemaking process within the United States

The rulemaking process in the United States includes an opportunity for public comment in between a new regulation and implementation of a rule change. During this comment period, not only the public at large, but corporations as well are able to exert influence over rule changes. Using textual analysis techniques, we aim to understand how this process of influence works – to what extent corporate influence affects rule changes and what kinds of changes ultimately result.

Mentor: Marianne Bertrand, Chris P. Dialynas Distinguished Service Professor of Economics and Willard Graham Faculty Scholar, Booth School of Business

Marianne Bertrand is the Chris P. Dialynas Distinguished Service Professor of Economics at the University of Chicago Booth School of Business. She is a Research Fellow at the National Bureau of Economic Research, the Center for Economic Policy Research, and the Institute for the Study of Labor.

Professor Bertrand is an applied micro-economist whose research covers the fields of labor economics, corporate finance, and development economics. Her research in these areas has been published widely, including numerous research articles in the Quarterly Journal of Economics, the Journal of Political Economy, the American Economic Review, and the Journal of Finance.

Professor Bertrand is Faculty Director of Chicago Booth’s Rustandy Center for Social Sector Innovation and the Faculty Director of the Poverty Lab at the University of Chicago Urban Labs. Professor Bertrand also serves as co-editor of the American Economic Review.

She has received several awards and honors, including the 2004 Elaine Bennett Research Prize, awarded by the American Economic Association to recognize and honor outstanding research in any field of economics by a woman at the beginning of her career, and the 2012 Society of Labor Economists’ Rosen Prize for Outstanding Contributions to Labor Economics. She is a Fellow of the American Academy of Arts and Sciences.

Born in Belgium, Professor Bertrand received a Bachelor’s Degree in economics from Belgium’s Universite Libre de Bruxelles in 1991, followed by a Master’s Degree in econometrics from the same institution the next year. She moved to the United States in 1993 and earned a Ph.D. in economics from Harvard University in 1998. She was a faculty member in the Department of Economics at Princeton University for two years before joining Chicago Booth in 2000.

Project: Combining human and machine intelligence for policy impact

The success of artificial intelligence (AI) for engineering and commercial applications has led to growing interest in using these tools to help solve important social problems like inequality in income, education, health, or criminal justice system involvement. But any realistic assessment of how AI will be used in these areas suggests it will be a complement to, not substitute for, human judgment. That is, AI will be used as decision aids, not decision makers. In previous work (Kleinberg, Lakkaraju, Leskovec, Ludwig and Mullainathan, 2018 Quarterly Journal of Economics) we have found in the context of criminal justice system decision-making that humans on net add negative value to the machine’s predictions of defendant risk, although in principle the private information humans can access that algorithms cannot (such as courtroom discussion about the details of the case) could help the human add positive value in at least some cases. Similar issues arise in numerous other policy domains such as medical diagnosis, hiring, credit, and education admissions. The goal of this project is to better understand the potential sources of human and machine comparative advantage by measuring the private information humans have access to in different decision-making domains, trying to understand what are useful sources of signal versus sources of noise for human decisions about when to follow versus over-ride the algorithm’s recommendations, and then try to build decision-making systems that lead to the human plus machine together to outperform the decisions implied by the machine’s predictions alone.

Mentor: Jens Ludwig, Edwin A. and Betty L. Bergman Distinguished Service Professor, Harris School of Public Policy, Director of University of Chicago Crime Lab, Co-director Education Lab

Jens Ludwig is the Edwin A. and Betty L. Bergman Distinguished Service Professor, director of the University of Chicago’s Crime Lab, codirector of the Education Lab, and codirector of the National Bureau of Economic Research’s working group on the economics of crime.

In the area of urban poverty, Ludwig has participated since 1995 on the evaluation of a HUD-funded randomized residential-mobility experiment known as Moving to Opportunity (MTO), which provides low-income public housing families the opportunity to relocate to private-market housing in less disadvantaged neighborhoods. In the area of education he has written extensively about early childhood interventions, and about the role of social conditions in affecting children’s schooling outcomes. In the area of crime, Ludwig has written extensively about gun-violence prevention. Through the Crime Lab he is also involved in partnering with policymakers in Chicago, New York City, and across the country to use tools from social science, behavioral science, and computer science to identify effective (and cost-effective) ways to help prevent crime and violence. This includes studies of various social programs, helping the Chicago Police Department use data to reduce gun violence and strengthen police-community relations, and work underway to use data science to help New York City build and implement a new pretrial risk tool as part of the city’s goal to close Riker’s Island. Crime Lab projects have helped redirect millions of dollars of public-sector resources to evidence-based strategies and have been featured in national news outlets such as the New York Times, Washington Post, Wall Street Journal, PBS News Hour and National Public Radio. In 2014 the Crime Lab was the recipient of a $1 million MacArthur Award for Creative and Effective Institutions, the organizational equivalent of the foundation’s “genius prize.”

His research has been published in leading scientific journals across a range of disciplines including Science, New England Journal of Medicine, Journal of the American Medical Association, American Economic Review, Quarterly Journal of Economics, the Economic Journal, and the American Journal of Sociology. His coauthored article on race, peer norms, and education with Philip Cook was awarded the Vernon Prize for the best article in the Journal of Policy Analysis and Management. He is also coauthor with Cook of Gun Violence: The Real Costs (Oxford University Press, 2000), coeditor with Cook of Evaluating Gun Policy (Brookings Institution Press, 2003), and coeditor with Cook and Justin McCray of Controlling Crime: Strategies and Tradeoffs (University of Chicago Press, 2012).

Prior to coming to Harris, Ludwig was a professor of public policy at Georgetown University. He is currently on the editorial board of the American Economic Review and was formerly coeditor of the Journal of Human Resources, and currently serves on the National Academy of Sciences Committee on the Neurobiological and Socio-behavioral Science of Adolescent Development and Its Applications. In 2012 he was elected vice president of the Association for Public Policy Analysis and Management (APPAM), the professional society for public policy schools. Ludwig received his BA in economics from Rutgers College and his MA and PhD in economics from Duke University. In 2006 he was awarded APPAM’s David N. Kershaw Prize for Contributions to Public Policy by Age 40. In 2012 he was elected to the Institute of Medicine of the National Academies of Science.

Project: Preventing violent encounters with first responders

People who live with serious mental illness or related challenges face heightened risks of violent encounters with first responders. Administrative and qualitative data from police, fire, and other first responders allow us to identify individuals, places, and events associated with such violent encounters. This project will use predictive analytics to improve preventive services and emergency responses for individuals and families who face these risks.

Project: Predicting mortality among high-users of safety-net services in Illinois

Individuals who pass through jails, homeless services, and other safety-net institutions face severe risks of premature mortality from opioid overdose, homicide, and other causes. This machine learning project uses integrated administrative data from diverse city, county, and state data sources in Illinois to identify key risk-factors for premature mortality.

Mentor: Harold Pollack, Helen Ross Professor of Social Service Administration, Co-director, University of Chicago Health Lab

Harold Pollack is the Helen Ross Professor at the School of Social Service Administration. He is also an Affiliate Professor in the Biological Sciences Collegiate Division and the Department of Public Health Sciences.

Co-founder of the University of Chicago Crime Lab, he is co-director of the University of Chicago Health Lab. He is a committee member of the Center for Health Administration Studies (CHAS) at the University of Chicago. His current NIH-funded research concerns improved services for individuals at the boundaries of the behavioral health and criminal justice systems, disabilities, and two major new efforts to address the opioid epidemic in Illinois and across the nation.

Past President of the Health Politics and Policy section of the American Political Science Association, Professor Pollack has been appointed to three committees of the National Academy of Sciences. He received his undergraduate degree, magna cum laude, in Electrical Engineering and Computer Science from Princeton University. He holds master’s and doctorate degrees in Public Policy from the Kennedy School of Government, Harvard University. Before coming to SSA, Professor Pollack was a Robert Wood Johnson Foundation Scholar in Health Policy Research at Yale University and taught Health Management and Policy at the University of Michigan School of Public Health.

He has published widely at the interface between poverty policy and public health. His research appears in such journals as Addiction, Journal of the American Medical Association, American Journal of Public Health, Health Services Research, Pediatrics, and Social Service Review. His journalism regularly appears in such outlets as Washington Post, the Nation, the New York Times, New Republic, and other popular publications. His American Prospect essay, “Lessons from an Emergency Room Nightmare” was selected for the collection Best American Medical Writing, 2009.

Project: Understanding the Effects of Gender on Policy Through Textual Analysis of Congressional Data

Congress generates a substantial amount of textual data from its hearings, meetings, speeches, etc. By conducting sentiment analysis on this data, we are trying to understand how gender influences the likelihood of member participation and the resulting policy decisions. Currently, this project has preliminary results and is entering a second phase of analysis, which will involve additional scraping of data and the generation of new methodologies for textual analysis.

Mentor: Heather Sarsons, Assistant Professor of Economics and Diane Swonk Faculty Fellow, Booth School of Business

Heather Sarsons is an economist with research interests in labor, personnel, and behavioral economics. Much of her work focuses on understanding how norms, stereotypes, and biases influence labor market outcomes and inequality.

Prior to joining Booth, Sarsons was a post-doctoral fellow at the University of Toronto’s GATE Institute and the U of T Economics Department.

Sarsons received a PhD in economics from Harvard, and a BA in economics from The University of British Columbia. While pursuing her PhD, she was also a visiting student at the London School of Economics.

Chenhao Tan is an assistant professor at the Department of Computer Science and the UChicago Data Science Institute. His main research interests include language and social dynamics, human-centered machine learning, and multi-community engagement. He is also broadly interested in computational social science, natural language processing, and artificial intelligence.

Homepage.

Project: News-based Sentiment Analysis to Understand Market Movement

Most financial analysis is quantitative, but there is a wealth of data contained in textual artifacts as well. In this project, we are using natural language processing methodologies to conduct sentiment analysis on global news reports in order to better understand how news-based sentiment affects the movement of markets. Other possible avenues of investigation using new NLP methodologies include understanding the macro effects of global economic sentiment and attempting to detect the existence of fake news.

Mentor: Dacheng Xiu, Professor of Econometrics and Statistics, Booth School of Business

Dacheng Xiu’s research interests include developing statistical methodologies and applying them to financial data, while exploring their economic implications. His earlier research involved risk measurement and portfolio management with high-frequency data and econometric modeling of derivatives. His current work focuses on developing machine learning solutions to big-data problems in empirical asset pricing.

Xiu’s work has appeared in Econometrica, the Journal of Econometrics, the Journal of the American Statistical Association, the Annals of Statistics, and the Journal of Finance. He is a Co-Editor for the Journal of Financial Econometrics, an Associate Editor for the Journal of Econometrics, the Journal of Business & Economic Statistics, the Journal of Empirical Finance, and Statistica Sinica, and also referees for several journals in the fields of econometrics, statistics, and finance. He has received several recognitions for his research, including the Fellow of the Society for Financial Econometrics, the Fellow of the Journal of Econometrics, the 2018 Swiss Finance Institute Outstanding Paper Award, the 2018 AQR Insight Award, and the Best Conference Paper Prize at the 2017 Annual Meeting of the European Finance Association.

In 2017, Xiu launched a website that provides up-to-date realized volatilities of individual stocks, as well as equity, currency, and commodity futures. These daily volatilities are calculated from the intraday transactions and the methodologies are based on his research of high-frequency data.

Xiu earned his PhD and MA in applied mathematics from Princeton University, where he was also a student at the Bendheim Center for Finance. Prior to his graduate studies, he obtained a BS in mathematics from the University of Science and Technology of China.

I’m a Postdoctoral Associate at the Hariri Institute for Computing at Boston University and the Boston University School of Law. I’ll be joining the University of Chicago as an Assistant Professor of Computer Science and Data Science starting January 2022.

My research explores the interplay between theoretical cryptography, privacy, law, and policy. Specifically, I aim to understand and resolve the tensions between the theory of cryptography and the privacy and surveillance law that governs its eventual real-world context. I earned my PhD in the Cryptography and Information Security group at MIT, advised by Shafi Goldwasser. Right now, I’m thinking about differential privacy, GDPR, the Fifth Amendment, encryption, multiparty computation, and the Census. Formerly an Affiliate at the Berkman Klein Center for Internet & Society, a Fellow at Facebook and at the Aspen Tech Policy Hub, and recipient of the NSF Graduate Student Research Fellowship.

I completed my PhD in Statistics in June 2020 at Stanford University, where I was lucky to be supervised by Professor Susan Holmes and where I had the chance to work with Prof. Jure Leskovec. Prior to Stanford, I studied Applied Mathematics and Engineering at Ecole Polytechnique (France), where I received an M.S and B.S equivalent.

My research interests lie in the analysis of patterns and the quantification of uncertainty in high-dimensional datasets, and in particular, graphs and networks, geared towards biomedical applications.

https://donnate.github.io/

Nick Feamster is Neubauer Professor in the Department of Computer Science and the College. He researches computer networking and networked systems, with a particular interest in Internet censorship, privacy, and the Internet of Things. His work on experimental networked systems and security aims to make networks easier to manage, more secure, and more available.

Homepage

I’m an assistant professor at the Department of Statistics at University of Chicago. I am also a member of Committee on Computational and Applied Mathematics (CCAM). I am interested in computational problems in structural biology and quantum many-body physics. I was extremely fortunate to have Amit Singer at Princeton as my Ph.D. adviser during 2012-2016, Lexing Ying at Stanford as my post-doc mentor during 2016-2019, and Phuan Ong at Princeton supervising my master thesis in experimental physics during 2010-2012.

https://www.stat.uchicago.edu/~ykhoo/

Sanjay Krishnan is an Assistant Professor of Computer Science. His research group studies the theory and practice of building decision systems that are robust to corrupted, missing, or otherwise uncertain data. His research brings together ideas from statistics/machine learning and database systems. His research group is currently studying systems that can analyze large amounts of video, certifiable accuracy guarantees in partially complete databases, and theoretical lower-bounds for lossy compression in relational databases.

Homepage.

My main research interests are in speech and language processing, as well as related aspects of machine learning.

I am an Associate Professor at TTI-Chicago, a philanthropically endowed academic computer science institute located on the University of Chicago campus. We are recruiting students to our PhD program and visiting student program, as well as additional faculty, including in speech and language-related areas (more on Speech and Language at TTIC).

I completed my PhD in 2005 at MIT in the Spoken Language Systems group of the Computer Science and Artificial Intelligence Laboratory. In 2005-2007 I was a post-doctoral lecturer in the MIT EECS department. In Feb.-Aug. 2008 I was a Research Assistant Professor at TTI-Chicago.

Homepage

David Uminsky joined the University of Chicago in September 2020 as a senior research associate and Executive Director of Data Science. He was previously an associate professor of Mathematics and Executive Director of the Data Institute at University of San Francisco (USF). His research interests are in machine learning, signal processing, pattern formation, and dynamical systems.  David is an associate editor of the Harvard Data Science Review.  He was selected in 2015 by the National Academy of Sciences as a Kavli Frontiers of Science Fellow. He is also the founding Director of the BS in Data Science at USF and served as Director of the MS in Data Science program from 2014-2019. During the summer of 2018, David served as the Director of Research for the Mathematical Science Research Institute Undergrad Program on the topic of Mathematical Data Science.

Before joining USF he was a combined NSF and UC President’s Fellow at UCLA, where he was awarded the Chancellor’s Award for outstanding postdoctoral research. He holds a Ph.D. in Mathematics from Boston University and a BS in Mathematics from Harvey Mudd College.

Rebecca Willett is a Professor of Statistics and Computer Science at the University of Chicago. She completed her PhD in Electrical and Computer Engineering at Rice University in 2005 and was an Assistant then tenured Associate Professor of Electrical and Computer Engineering at Duke University from 2005 to 2013. She was an Associate Professor of Electrical and Computer Engineering, Harvey D. Spangler Faculty Scholar, and Fellow of the Wisconsin Institutes for Discovery at the University of Wisconsin-Madison from 2013 to 2018.  Prof. Willett received the National Science Foundation CAREER Award in 2007, was a member of the DARPA Computer Science Study Group 2007-2011, and received an Air Force Office of Scientific Research Young Investigator Program award in 2010. Prof. Willett has also held visiting researcher positions at the Institute for Pure and Applied Mathematics at UCLA in 2004, the University of Wisconsin-Madison 2003-2005, the French National Institute for Research in Computer Science and Control (INRIA) in 2003, and the Applied Science Research and Development Laboratory at GE Medical Systems (now GE Healthcare) in 2002. Her research interests include network and imaging science with applications in medical imaging, wireless sensor networks, astronomy, and social networks. She is also an instructor for FEMMES (Females Excelling More in Math Engineering and Science; news article here) and a local exhibit leader for Sally Ride Festivals. She was a recipient of the National Science Foundation Graduate Research Fellowship, the Rice University Presidential Scholarship, the Society of Women Engineers Caterpillar Scholarship, and the Angier B. Duke Memorial Scholarship.

Homepage