This document summarizes a project that aimed to predict restaurant ratings and changes in popularity on Yelp using machine learning algorithms. The project used data from Yelp, including reviews, ratings, and text to train logistic regression, naive Bayes, and multinomial naive Bayes models. While logistic regression performed best, all methods' predictions were imperfect, indicating room for improvement through additional data or methods. The goals were to forecast ratings and popularity shifts based on restaurant features.