Bayesian hierarchical model in STAN using a room-type top level to predict listing price for Airbnbs in Toronto.
COVID-19 US Vaccination Modeling
Binomial regression model predicting proportion of the population aged 18+ who are fully vaccinated by US county.
Latent Dirichlet Allocation (LDA) Topic Modeling
LDA topic model using Gensim on Department of Justice press release data (2009-2018).
Compared different classifiers such as decision tree, random forest, boosting tree and K-NN to predict debt crisis events. Used k-fold cross-validation for hyperparameter tuning. Examined AUC score and recall as measures of accuracy.