Monday, Aug 7: 2:00 PM - 3:50 PM
Invited Paper Session
Metro Toronto Convention Centre
Justice Equity Diversity and Inclusion Outreach Group
Caucus for Women in Statistics
Committee on Minorities in Statistics
Throughout the last decade, an underlying theme in my work has been communicating technical information to broad audiences. As part of my biostatistics PhD program, I investigated ways to use nutritional information from the Multi-Ethnic Study of Atherosclerosis to identify heart-healthy diet patterns. I evaluated ways to summarize the diet data using partial least squares and sparsity techniques, and how to interpret and visualize patterns in the data that were associated with a participant's first heart disease event. As the pandemic developed, I joined a cross-institutional student-led group to clarify COVID-19 guidance for people in Mexico through news articles. After graduation, I served as a Science Policy Fellow, training for 2 months about how laws and decisions get made in California before being placed for the rest of the year in the Department of Pesticide Regulation. I collaborated on materials for public webinars and internal briefing documents, highlighting progress on scientific and policy projects in the context of government processes. In this talk, I will share some lessons learned that have enabled me to adapt between research, science communication, and policy.
Over the past 3 years, data has been collected on high and middle school students that participate in GATO365 Data Science Academy, an educational curriculum development company founded by Immanuel Williams. The GATO365 curriculum is designed to adjust based on students' abilities and questions, incorporating Dr. Williams' research on culturally relevant data, extraction, transformation, visualization of data and the implementation of a deeper knowledge of APIs in the data science classroom. This data-driven approach will help instructors and curriculum developers determine what material should be taught in middle school, high school, and college to optimize data science education. This presentation will discuss the adaptive curriculum and assessments used to navigate and enhance data science education for students at various levels.
Dynamic treatment regimes formalize precision medicine as a sequence of decision rules, one for each stage of clinical intervention, that map current patient information to a recommended intervention. Optimal regimes are typically defined as maximizing some functional of a scalar outcome's distribution, e.g., the distribution's mean or median. However, in many clinical applications, there are multiple outcomes of interest that are not easy combined into a single scalar outcome. We consider the problem of estimating an optimal regime when there are multiple outcomes ordered by priority but which cannot be readily combined by domain experts into a single scalar outcome. We propose a definition of optimality in this setting and show that this definition leads to maximal mean utility under a large class of utility functions. Furthermore, we use inverse reinforcement learning to identify a composite outcome that most closely aligns with our definition within a pre-specified class. Simulation experiments and an application to data from a sequential multiple assignment randomized trial (SMART) on HIV/STI prevention illustrate the usefulness of the proposed approach.
The Pacific Northwest National Laboratory (PNNL) is a U.S. Department of Energy Office of Science National Laboratory with core capabilities that span several scientific domains, including chemical and material science, engineering, biological and earth sciences, and computational and mathematical sciences. As such, PNNL data scientists enjoy a rich variety of interesting and challenging problems to choose from in providing their expertise. This presentation showcases this diversity by discussing two examples of PNNL project work: an application of Generalized Additive Models for Location Scale and Shape (GAMLSS) models towards estimating false discovery rates in gas-chromatography mass-spectrometry (GC-MS) metabolomics; and the development of a new, model-agnostic inferential test for detecting interaction effects.