Events
FDS Project Match
Tuesday, December 10, 2024
3:00PM - 4:00PM
Location: Yale Institute for Foundations of Data Science, Kline Tower 13th Floor, Room 1327, New Haven, CT 06511
Additionally: Remote access available via Webcast
Add To: Google Calendar | Outlook | iCal File
The FDS Data Science Project Match, hosted by the Yale Institute for Foundations of Data Science (FDS), is an opportunity for Yale faculty from any department or school within the university to connect with talented students from the departments of Statistics and Data Science, Applied Mathematics, and Computer Science. In a series of lightning-round talks, faculty will have exactly five minutes to pitch a current research problem, aiming to team up with students interested in tackling complex data challenges. This event facilitates collaboration on current research projects, offering a platform for faculty to present their data-driven initiatives and find skilled undergraduate and/or graduate students eager to contribute. It’s also a wonderful way to learn about the research of many Yale faculty.
Presenters:
- Jennifer Marlon
Senior Research Scientist, Lecturer, and Director of Data Science,
Yale School of the Environment
Executive Director, Yale Center for Geospatial Solutions
Presented by Andrew Gillreath-Brown
jennifer.marlon@yale.edu“Aligning Public Perception with Environmental Hazards (YPCCC)“
This project investigates the gap between public worry about climate-driven hazards like wildfire, drought, and extreme heat and the actual vulnerability levels faced by communities across the United States. By integrating nationally representative survey data on public perception with objective measures of risk—such as wildfire indices, drought severity scores, and heat vulnerability metrics derived from emergency room visits and environmental data—you’ll help uncover where perception misaligns with reality. Your work will directly contribute to tailoring risk communication, improving resource allocation, and building climate resilience. This is a unique opportunity to apply your technical skills in computer science, data science, or applied math to cutting-edge causal inference methods while gaining experience with large-scale survey, environmental, and health datasets. Join us to help bridge the gap between perception and reality, empowering communities and policymakers to act effectively in the face of climate change.
“Geospatial Analysis and Earth Observation (YCGS)“
The Yale Center for Geographic Solutions (YCGS) is seeking an enthusiastic student to join our research team focused on applying Earth observation (EO) data and geospatial analysis to tackle pressing global challenges. The exact project will be determined based on ongoing research priorities and the student’s skills and interests, but all projects will be conducted in collaboration with YCGS-affiliated faculty, staff, and students, and will provide valuable experience with state-of-the-art geospatial tools and methods.
Examples of potential projects include: 1) Enhanced Flood Zone Mapping in Connecticut: Work with high-resolution LiDAR data to map terrain, vegetation, and urban infrastructure across Connecticut, and integrate this data with satellite imagery and climate datasets to analyze environmental patterns and contribute to projects addressing flood risk, urban heat islands, and land cover change; 2) Assessing Infrastructure Development in Developing Countries: Use satellite imagery and geospatial datasets to monitor infrastructure construction progress. Collect and preprocess project data (e.g., timelines, locations, descriptions) using web scraping techniques, and integrate these data with EO data to evaluate policy effectiveness and project outcomes; and 3) Urban Verticality and Social Equity: Integrate a groundbreaking global building height dataset derived from remote sensing with rich social science data to examine whether areas of high-density vertical growth are more or less resilient to climate challenges, such as heat waves and air pollution. In all cases, the student will gain hands-on experience with advanced tools like Google Earth Engine, Python, and GIS platforms (e.g., QGIS) for data processing, analysis, and visualization. They will also learn techniques for integrating diverse data sources, including LiDAR, satellite imagery, and vector data, to address real-world challenges such as climate resilience, environmental conservation, and sustainable development.
- Kartik Pattabiraman
Assistant Professor, Child Study Center, Yale School of Medicine
kartik.pattabiraman@yale.edu | https://medicine.yale.edu/profile/kartik-pattabiraman/“Creating high throughput approaches to causally link psychiatric risk genes to circuit vulnerabilities”
Neuropsychiatric disorders are currently the leading cause of disability in the United States. Limited understanding of the underlying pathophysiology of these disorders has resulted in non-specific pharmacologic interventions with minimal to moderate efficacy. Patient-based studies combining the fields of psychiatry, neuroimaging, and genetics have provided substantial evidence for a shared neurodevelopmental etiology. Translating these research findings into effective outcome-altering treatments requires establishing direct causal links between the genetic, neuroimaging, and behavioral signatures of psychiatric disorders. However, the current strategies used to study brain connectivity are too laborious to characterize the thousands of identified risk alleles and lack the resolution required to identify the single circuit-level dysconnectivity underlying these disorders. This project aims to create high throughput and high-resolution strategies to directly link individual risk alleles to changes in brain connectivity using automated characterization of direct fluorescent labeling of neuronal projections from tissue-cleared mouse lines.
- Johan Ugander
Associate Professor at Stanford University in the Department of Management Science & Engineering and the Institute for Computational & Mathematical Engineering
Visiting faculty at FDS
johan.ugander@yale.edu | https://stanford.edu/~jugander/“Strategic behavior with bridging-based fact-checking”
Social networks scaffold the diffusion of information on social media, both true and false. A recent fact-checking feature adopted by Twitter (now X) asks volunteers to propose and vet fact-checking notes for false content. An important aspect of this feature, called Community Notes, is its use of a bridging-based decision algorithm whereby fact-checking notes are shown only if they are seen as broadly informative and helpful by users from across the political spectrum. This design is understood to prevent partisan engagement. The goal of this proposal is to investigate strategic misuse of the Community Notes feature by partisan actors.
- Karen Seto
Frederick C. Hixon Professor of Geography & Urbanization Science
Faculty Director, Hixon Center for Urban Sustainability
Co-Director, Yale Center for Geospatial Solutions
U.S. National Academy of Sciences Council on Foreign Relations
Yale School of the Environment
Karen.seto@yale.edu | http://urbanization.yale.edu
Presented by Juwon Kong, Postdoctoral Fellow, YSE“Fusing multi-resolution, multi-sensor satellite data to explore environmental change”
This project uses satellite imagery with different spatial, spectral, temporal and radiometric resolutions from multiple sensors and fuses them into a single high-resolution daily product to capture decades of environmental change in parks across the US. Specifically, we use satellite imagery from the MODIS, Landsat, and Planet sensors to create a coherent dataset that leverages the benefits of each individual dataset. The RA will explore different models to fuse satellite imagery of varying resolutions, including pyramid-based methods, wavelet transform-based methods, statistical approaches, and deep learning methods.
- Leandros Tassiulas
John C. Malone Professor of Electrical & Computer Engineering & Computer Science, Yale School of Engineering and Applied Science
Presented by Ali Maatook, Postdoc in the Tassiulas Lab
leandros.tassiulas@yale.edu |“LLM customization”
- Kiran Turaga
Professor of Surgery, Chief of Surgical Oncology, Yale School of Medicine
Assistant Director of CTO, Yale Cancer Center
kiran.turaga@yale.edu | https://medicine.yale.edu/profile/kiran-turaga/“Registry based clinical trials”
Clinical trials are the cornerstone in determining new evidence for healthcare but are incredibly expensive and difficult to perform. We propose a novel mechanism for discovery of therapeutics and conduct of clinical trials.
- Kristen Brennand
Elizabeth Mears and House Jameson Professor of Psychiatry
Co-Director, Yale Science Fellows Program
Yale School of Medicine
kristen.brennand@yale.edu | https://brennandlab.org“Using Stem Cells to Explore the Genetics Underlying Brain Disease”
Each person’s distinct genetics and environment predispose them to some phenotypes and confers resilience to others. How do all the individual variants across the genetic landscape combine to yield larger phenotypic impacts in aggregate? How does genetic variation govern the penetrance of deleterious mutations, variable expressivity, and pleiotropy? What is the role of the environment across the lifespan? Understanding how these elements interact will advance our knowledge of human development, aging, health, and disease. Our functional genomics approach integrates human induced pluripotent stem cell models with CRISPR-based genome engineering to introduce and reverse genetic variation, yielding precision models that can be combined with genetic and pharmacological screens. With this approach, we demonstrated that diverse risk variants share downstream convergent impacts, and that when added together, their combinatorial perturbations yield novel non-additive outcomes that cannot yet be predicted by individual manipulations alone. We seek to understand the genetic regulation of phenotype, and how it is impacted by developmental, cellular, and environmental contexts. Thus, rather than just characterize the impact of trait-associated variants, we seek to uncover modifiers that alter it. For example, we study how genotype-phenotype relationships vary across people and dynamic conditions. Our goal is to decipher the frameworks that buffer genetic risk, in order to confer biological resilience and promote healthy development. We are uniquely positioned to answer critical questions: How does the environment impact genetic regulation? Why are there marked sex effects across many human traits and diseases? What are the molecular mechanisms of resilience, whereby individuals with high genetic risk show no clinical manifestation of disease? Understanding the basic biology governing the complex interplay between genetic variants and the environment will springboard the development of novel, personalized approaches to improve health and prevent disease.
- Purushottam Dixit
Assistant Professor, Department of Biomedical Engineering,
Yale School of Engineering and Applied Science
purushottam.dixit@yale.edu | https://sites.google.com/view/dixitlab“Mapping the Waddington landscape of mammalian tissues using Hopfield networks”
In higher organisms, different tissue types, e.g., eyes, the blood, and the brain originate from the same DNA. How the same instruction manual (DNA) is used to create very different machines (tissues) is an open problem in organismal biology. In the 1950s, Waddington provided a qualitative picture of the process of canalization to form tissues using the language of dynamical systems. However, till recently, this remained a qualitative picture. Recent advances in sequencing allow us to collect large amounts of high dimensional gene expression data across multiple tissues. We propose a project to use this data and Hopfield networks to construct the Waddington landscape of mammalian tissue formation.
Add To: Google Calendar | Outlook | iCal File
Submit an Event
Interested in creating your own event, or have an event to share? Please fill the form if you’d like to send us an event you’d like to have added to the calendar.