History of median home value fluctuation over time with several other factors: 1) housing volume (stock), 2) census data estimated population change, 3) mortgage rates. Housing data taken from Zillow research, census data taken from US Census Bureau, mortgage rates taken from Federal Reserve Economic Data.

Tables showing number of cases, case fatality rate and time to death: break down by age group and month

A dotplot of COVID-19 deaths broken down by age (two groups with age 60 as cutoff) and county. A choropleth of Ohio counties with shading indicating percentage of county that is 60+ in age.

Maps showing county-by-county break down of: 1) COVID-19 case totals 2) COVID-19 prevalence 3) COVID-19 death rates. Main story centers on COVID-19 outbreaks in two prisons.

Use of Hastie's elastic net algorithm with alpha = 0 for ridge regression. This ridge model is comparable to LASSO in terms of accuracy and sensitivity with predicting Parkinson's disease from audio recordings.

A brief bio of myself.

Demonstration of LASSO for feature selection to enhance prediction performance.

Logistic models are used with probability cuts optimized through a grid search. Basic feature selection approaches are used. Results in ~64% accuracy of predicting Parkinson's (with 74% sensitivity).