/Delhi_Rental_Price_Prediction

End-End Data Science project. Using Scraped rental listings and an Ensemble Voting Regressor to predict monthly rent.

Primary LanguageJupyter Notebook

Delhi Rental Price Prediction

Estimate the rent you should be paying for a house in New Delhi. Models trained on 12000 scrapped public rental listings and deployed as a Flask app.

  1. Data scraping and ingestion scripts.
  2. Data Cleaning and Wrangling Notebook.
  3. Analysis and Modelling- Using a Voting Regressor Ensemble of Random Forest and XGBoost with one model optimised for reduced MAE and the other for reduced RMSE.
    All Feature selection (RFE), model comparisons and hyper-parameter optimisation in Notebook.
  4. Deployed as a Flask App.
Model Pipeline Web-App
models App