The promise of machine learning for responding to epidemics and other public health crises
Session Co-chair: Magdalena Cerda, New York University
Session Co-chair: Brandon Marshall, Brown university
Machine learning approaches are increasingly applied in epidemiology, yet the specific contributions they can make to addressing important public health questions are still not well understood. This symposium will illustrate the use of machine learning approaches to forecast risk and implement mitigation strategies during public health crises. Presentations will discuss the use of machine learning to target interventions at the individual, neighborhood, and state levels in an effective and equitable manner. Topics include: an assessment of the contributions of nonparametric estimation approaches using machine learning to estimate causal effects; using clinical decision support with an embedded prediction model to improve PrEP prescribing in diverse healthcare settings; using machine learning to identify community predictors of future overdose risk; balancing spatial equity versus impact when predicting future hot spots to target overdose prevention efforts; and deep learning data-driven approaches to forecast the progression of the COVID-19 pandemic. Through these presentations, we hope to illustrate the potential that machine learning tools offer to inform the design, implementation, and evaluation of public health strategies.
Presenters:
Kara Rudolph, Columbia University
“All models are wrong, but does it matter? On the bias of parametric estimators of causal effects in randomized and observational studies.”
Julia Marcus, Harvard University
“Using predictive analytics to improve PrEP prescribing in diverse healthcare settings”
Robert Schell, University of California, Berkeley
“Using Machine Learning to Identify the Community-Level Predictors of Opioid Overdose Deaths”
Bennett Allen, New York University
“Machine learning to target overdose prevention: A modeling strategy and evaluative framework for public health practice”
B. Aditya Prakash, Georgia Institute of Technology
“Deep Learning data-driven approaches for Epidemic Forecasting”