In this project, we focused on predicting Yelp 5-star rating based on the review text, and other user-related information. We considered this as a regression problem, and compared the results for various methods. We built 3 (Ordinary Least Squares, Lasso, Polynomial Regression) different prediction models using 9 predictors and analyzed the performance of each model to find the best model for predicting.
The complete story is described in the pdf file. Code implementation can be found in case2.ipynb. Data source can be found within the notebook.