Time-Varying Effects of Meals and Insulin on Postprandial Glucose Response Using Autoencoder-Based Causal Representations
Presented autoencoder-based causal mediation methods for postprandial glucose response ahead of the ENAR Spring Meeting.
Professional Summary
Spencer Hilligoss is a PhD candidate in Statistics at the University of California, Irvine,
advised by Dr. Annie Qu and Dr. Tianchen Qian. His research develops methods for causal
mediation analysis, representation learning, and generative modeling of complex time series,
with applications spanning personalized medicine, mobile health, and financial data. His
work on causal mediation in continuous glucose monitoring was recently published in
npj Metabolic Health and Disease.
Education
PhD Statistics
2023-09-01
University of California, Irvine
BS Statistics and Operations Research
2019-08-01
2023-05-14
University of North Carolina at Chapel Hill
BA Economics
2019-08-01
2023-05-14
University of North Carolina at Chapel Hill
Interests
Representation Learning
Statistical Machine Learning
Personalized Medicine
Mobile Health
Causal Inference and Mediation Analysis
Research
My research focuses on developing statistical methods for healthcare applications, particularly in the context of mobile health and personalized medicine. I am currently working on two main projects:
Causal Mediation Analysis for Type 1 Diabetes — Developing autoencoder-based methods to understand how meal carbohydrate intake affects post-meal glucose trajectories through insulin bolusing behavior.
Tensor-Based Reinforcement Learning for Adaptive Insulin Dosing — Building personalized treatment recommendation frameworks using tensor factorization methods for longitudinal continuous glucose monitoring data.
Publications
(2026).
Causal Mediation Pathways in Continuous Postprandial Glucose Monitoring for Type 1 Diabetes Patients.
npj Metabolic Health and Disease 4, 35 (2026).
(2026).
Selective Inference for Time-Varying Causal Effect Moderation via the Ordered Lasso.
Manuscript in preparation.
Presentations
Time-Varying Effects of Meals and Insulin on Postprandial Glucose Response Using Autoencoder-Based Causal Representations
Presented research on autoencoder-based causal mediation methods for analyzing postprandial glucose response in Type 1 Diabetes.
Teaching
Guest Lecturer: STATS 295 - Special Topics in Machine Learning
Guest lecturer for STATS 295, a graduate-level special topics course in machine learning.
Mentorship
Bryson Harris (UC Santa Barbara) — Mentoring on financial time series modeling through weekly research meetings.
Daniel You (UC Santa Barbara) — Mentoring on generative diffusion modeling with an application to soccer tactics through weekly research meetings.
Awards & Fellowships
T32 STEER in Biomedical Sciences Fellow (2025, renewed 2026) — NIH-funded training fellowship supporting research at the intersection of statistics and biomedical sciences, University of California, Irvine.
Diversity Recruitment Fellowship (2023) — Fellowship awarded to support doctoral studies in Statistics at the University of California, Irvine.
PREDOC Summer Program (2022) — Selected for competitive data analytics training program at Harvard University’s Opportunity Insights, directed by economists Raj Chetty, John Friedman, and Nathaniel Hendren.
Projects
Hidden Markov Model for Part-of-Speech Tagging Tasks
Literature review and application of Hidden Markov Model methods to the Penn Treebank dataset for STATS 230: Statistical Computing Methods. Tools Used: Python, R
•
1 min read
Predicting the Dow Jones Industrial Average with Sentiment-Enhanced LSTM Models
Project demonstrating the efficacy of LSTM models in enhancing prediction of stock indexes such as the DJIA. Tools Used: Python, R
•
1 min read
The Lasting Impact of HOLC Redlining on Minorities in New York
Short research paper completed as part of the PREDOC Summer Program 2022 at Harvard University’s Opportunity Insights, a data analytics training course directed by economists Raj …
•
1 min read
Contact
Feel free to reach out via email at shilligo@uci.edu.
Office: Department of Statistics, University of California, Irvine