Disease Cluster Identification @TNHSRP

Created a county level model to predict COVID-19 cases(on a daily basis) and Tuberculosis cases (on a yearly basis) separately. We considered various socio-economic, geographic and health-based factors with disease data, to simulate varied rates of growth. After removing features with low correlation, we used a Neural Network model which was trained using both Backpropogation and Genetic Algorithm for weights. We had also considered a Random Forest model which provided decreased results.