Beyond a Machine Learning Model
summary
Uses Johns Hopkins data and the CoronaNet dataset to train models that predict country-level COVID case growth rates and then interprets effects with ALE plots package built for the project. Finds school closures and bans on mass gatherings show the largest reduction in growth, with effects appearing about 10 days after implementation; national mandatory social distancing and business restrictions follow. Releases a Python ALE visualization library (1D/2D, categorical handling, confidence intervals) and publishes results in PMC Public Health.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:09Beyond predictions
- 3:19Cumulative local effects
- 5:05How CLE works
- 7:17COVID-19 case study
- 15:30Project outputs
- 16:42alePython package
- 19:34Q&A
related talks
22mCOVID-19: Data Science & Expertise
29mHow data science shortened the COVID-19 pandemic by 2 months
01h 05mBuilding a coronaradar
13mGeometry for Social Good
33mCOVIDcast: An ecosystem for COVID-19 tracking and forecasting
37mAI/ML Strategies & Techniques for a Post Pandemic World
50mReporting on and visualising the pandemic
58mKeynote: DJ Patil