Explore a 22-minute conference talk from GOTO Chicago 2020 that delves into data science applications for COVID-19 prediction. Learn about two projects: predicting county-level COVID cases in the United States and detecting COVID using chest x-rays. Gain insights into predictive modeling techniques, the importance of subject matter expertise in model validation, and the use of AutoML to lower barriers in AI implementation. Discover how socio-economic data is utilized for daily predictions, and understand the process of evaluating and explaining COVID-19 models. The presentation also covers the use of chest x-ray data and COVID embeddings in prediction efforts. Enhance your understanding of data science's role in addressing the global pandemic through real-world examples and expert analysis.
Overview
Syllabus
Intro
Machine learning models
Reality is more complex than one variable
Lowering the barrier to Al - AutoML
Expertise in Real Estate
Data Science and COVID-19
COVID-19 in the United States
Daily Predictions to identify Counties to Have Covid-19
Daily Predictions for Covid-19
Start with Socio-Economic Data
Evaluating the Model
Explaining the Model
Next Steps -- Second Generation
What data was used?
COVID-19 Chest X-rays
COVID Embeddings
Taught by
GOTO Conferences
