/hugo

A webapp for predicting restaurant health inspection results using Google Maps data

Primary LanguageJupyter Notebook

Hugo

A webapp for predicting restaurant health inspection results using Google Maps data

Created for The Data Incubator capstone project, fulfilling requirements:

  1. Clear business objective
  • The webapp allows users to obtain information about the sanitary conditions at food establishments which may not be posted in an easily-accesible format online or in the business.
  1. Data ingestion
  • Data was downloaded from the New York City Department of Health and Mental Hygiene as well as from the Google Maps API.
  1. Visualizations
  • Several visualizations illustrating model fits to the data are plotted in the model development Jupyter notebook model/Food Safety Model Development.ipynb.
  1. (a) Machine learning
  • The predictive model uses natural language processing, regression, and cross validation
  1. (c) Interactive website
  • The web app allows users to search for a restaurant through Google Maps. Data from the Maps API is then given to the predictive model to provide an estimate of the restaurant's health inspection performance.
  1. A deliverable
  • The Jupyter notebook model/Food Safety Model Development.ipynb details data processing and development of the model.

Notes:

  • The web app requires a working Google Maps API key, stored in API_keys/API_key_GoogleMaps.txt.
  • To run the webapp navigate to the 'webapp/' directory and run streamlit run hugo.py
  • Data used for model training is located in model/data/raw_data.zip and must be unzipped before running the model development Jupyter notebook.