Publications
Paper name: Towards Accurate Spatiotemporal COVID-19 Risk Scores using High Resolution Real-World Mobility Data
Description: The paper discusses the importance and a new method for accurately predicting COVID-19 risk scores through the use of fine-grained mobility data. To test this, the team uses a Hawkes process-based spatiotemporal risk measure "LocationRisk@T" to track infections on simulated disease spread mobility. This data is generated by an agent simulation "SpreadSim" on cell-phone location signals.
My Role: Implemented and helped design the referenced agent simulation "SpreadSim" in C++ to mimic the spread of disease, and used it to generate the ground-truth number of infections in major cities for the experiments using real world high-resolution mobility patterns.