The dataset was collected from Yelp Open Dataset at two different periods: 6/2/2018 and 5/23/2019, with each dataset updated by Yelp on 8/22/2017 and 12/19/2018 respectively. Therefore, we shall assume that the information contained in both datasets differ by about 1 year. The problem we study is predicting whether a restaurant will be bankruptcy or not in next year based on the restaurant features in Yelp Official Dataset.
Since the dataset is too large to be uploaded to UCSB class (INT 15 - Spring 2019) JupyterHub server, we have preprocessed the data by selecting a subset of the original dataset. The code for the data preprocessing steps can be found in Data_Preprocessing.ipynb. The master code is Yelp_Bankruptcy_Analysis.ipynb.
WARNING: The code usually need about 2 hours to run.