I am an Assistant Professor in the School for Environment and Sustainability at the University of Michigan and am part of the Sustainable Food Systems Initiative. My research examines the impacts of environmental change on agricultural production, and how farmers may adapt to reduce negative impacts. I also examine ways that we can sustainably enhance agricultural production. To do this work, I combine remote sensing and geospatial analyses with household-level and census datasets to examine farmer decision-making and agricultural production across large spatial and temporal scales.
Our research is focused on Post ICU pain syndromes (PIPS). PIPS exhibit distinct phenotypic presentations and can be predicted by intra-ICU parameters. Our primary goal is to be able to predict post-ICU opioid use based on intra-ICU parameters. We utilize a data-driven characterization of post-ICU pain syndromes will utilize unsupervised clustering algorithms including DBSCAN and spectral clustering. Prediction of post-discharge pain severity, likelihood of specific pain presentations, and post-discharge opioid use will be achieved using logistic LASSO, random forests, and neural networks. Specifically, these tests will utilize available ICU data to predict changes between pre-
and post-ICU pain severity, incidence of specific pain presentations, and incidence of opioid use.
Edgar Franco-Vivanco is an Assistant Professor of Political Science and a faculty associate at the Center for Political Studies. His research interests include Latin American politics, historical political economy, criminal violence, and indigenous politics.
Prof. Franco-Vivanco is interested in implementing machine learning tools to improve the analysis of historical data, in particular handwritten documents. He is also working in the application of text analysis to study indigenous languages. In a parallel research agenda, he explores how marginalized communities interact with criminal organizations and abusive policing in Latin America. As part of this research, he is using NLP tools to identify different types of criminal behavior.
Yixin Wang works in the fields of Bayesian statistics, machine learning, and causal inference, with applications to recommender systems, text data, and genetics. She also works on algorithmic fairness and reinforcement learning, often via connections to causality. Her research centers around developing practical and trustworthy machine learning algorithms for large datasets that can enhance scientific understandings and inform daily decision-making. Her research interests lie in the intersection of theory and applications.
Dr. VanEseltine is a sociologist and data scientist working with large-scale administrative data for causal and policy analysis. His interests include studying the effects of scientific infrastructure, training, and initiatives, as well as the development of open, sustainable, and replicable systems for data construction, curation, and dissemination. As part of the Institute for Research on Innovation and Science (IRIS), he contributes to record linkage and data improvements in the research community releases of UMETRICS, a data system built from integrated records on federal award funding and spending from dozens of American universities. Dr. VanEseltine’s recent work includes studying the impacts of COVID-19 on academic research activity.
My research focuses on building infrastructure for public health and health science research organizations to take advantage of cloud computing, strong software engineering practices, and MLOps (machine learning operations). By equipping biomedical research groups with tools that facilitate automation, better documentation, and portable code, we can improve the reproducibility and rigor of science while scaling up the kind of data collection and analysis possible.
Research topics include:
1. Open source software and cloud infrastructure for research,
2. Software development practices and conventions that work for academic units, like labs or research centers, and
3. The organizational factors that encourage best practices in reproducibility, data management, and transparency
The practice of science is a tug of war between competing incentives: the drive to do a lot fast, and the need to generate reproducible work. As data grows in size, code increases in complexity and the number of collaborators and institutions involved goes up, it becomes harder to preserve all the “artifacts” needed to understand and recreate your own work. Technical AND cultural solutions will be needed to keep data-centric research rigorous, shareable, and transparent to the broader scientific community.
View MIDAS Faculty Research Pitch, Fall 2021
My research interests include health effects of air pollution, temperature extremes and climate change (mortality, asthma, hospital admissions, birth outcomes and cardiovascular endpoints); environmental exposure assessment; and socio-economic influences on health.
Data science tools and methodologies include geographic information systems and spatio-temporal analysis, epidemiologic study design and data management.
As an environmental epidemiologist and in collaboration with government and community partners, I study how social, economic, health, and built environment characteristics and/or air quality affect vulnerability to extreme heat and extreme precipitation. This research will help cities understand how to adapt to heat, heat waves, higher pollen levels, and heavy rainfall in a changing climate.
The primary tools currently in use are variations of linear models (regression, MLM, SEM, and so on) as we pursue the initial aims of the NICHD funded work. We are expanding into new areas that require new tools. Our adolescent sample is diverse, selected through quota sampling of high schools close enough to UM to afford the use of neuroimaging tools, but it is not population representative. To overcome this, we have begun work to calibrate our sample with the nationally representative Monitoring the Future study, implementing pseudo-weighting and multilevel regression and post-stratification. To enable much more powerful analyses, we are aiming toward the harmonization of multiple, high quality longitudinal databases from adolescence through early adulthood. This would benefit traditional analyses by allowing cross-validation with high power, but also provide opportunities for newer data science tools such as computational modeling and machine learning approaches.