Project 5 - Predicting Yelp Ratings

This repo contains code for my third project at Metis, a data science bootcamp in NYC. For this project, I employed logistic regression using both sci-kit learn and maximum likelihood estimation to classify a business’ Yelp rating. The data was sourced from Yelp's Dataset Challenge. My ultimate goal was to create a tool to help illuminate the effect of Yelp-specific levers on ratings via a beta coefficient analysis.

yelp_data_exploration.ipynb: Contains code relating to data exploration, data pre-processing, and feature engineering.
yelp_rating_classification.ipynb: Contains code relating to feature selection, model selection, and visualization.