Tracking the Spatiotemporal Spread of the Ohio Overdose Epidemic with Topological Data Analysis
By: Nicholas Bermingham, David White, Nathan Willey
Potential Business Impact:
Finds drug overdose patterns to help stop them.
In recent years, techniques from Topological Data Analysis (TDA) have proven effective at capturing spatial features of multidimensional data. However, applying TDA to spatiotemporal data remains relatively underexplored. In this work, we extend previous studies of disease spread by using the Mapper algorithm to analyze the Ohio drug overdose epidemic from 2007 to 2024. We introduce a novel method for constructing covers in Mapper graphs of spatiotemporal data that respects geographic structure and highlights the time-dependent variables. Finally, we generate a Mapper visualization of regional demographics to examine how these factors relate to overdose deaths. Our approach effectively reveals temporal trends, overdose hotspots, and time-lagged patterns in relation to both geography and community demographics.
Similar Papers
Tracking Temporal Evolution of Topological Features in Image Data
Methodology
Finds patterns in changing pictures over time.
A Bayesian Spatio-Temporal Top-Down Framework for Estimating Opioid Use Disorder Risk Under Data Sparsity
Applications
Find drug use where data is missing.
Topological Data Analysis for Unsupervised Anomaly Detection and Customer Segmentation on Banking Data
Machine Learning (CS)
Finds hidden customer habits in bank data.