HDB Flat Resale Price Prediction

Overview of Project Components

The following are a brief explanation of the different folders, document and files in this project repository:

1. Report - Resale Price Prediction of HDB Resale Flats in Singapore

• This is the full write-up report for the project.
• The purpose of the report is to provide a succint update of the entire project to project stakeholders.

2. Notebook - HDB Flat Resale Price Prediction

• Jupyter Notebook that contains the full codes of the project and detailed explanation and comments for each step of the process.
• The notebook is intended for users and collaborators to access the full codes of the project.

3. Folder - Data

• This folder contains all the datasets used in this project.
• The various datasets are saved in this folder to reduce resources required to extract the full data from various APIs.
  • dat_hdb.csv: HDB Flat Resale Transaction Data (Jan 2017 - Jan 2020), retrieved from Data.gov.sg
  • dat_coord.csv: Coordinate data for each transaction address extracted via geocoding
  • dat_venues.csv: Nearby venues data for each transaction location within 500m radius, extracted from Foursquare
  • dat_results.csv: Summary table of the performance metrics between the two models
  • dat_pred.csv: Prediction results vs actual transaction price of HDB Resale Flats.
  • dat_final.csv: Processed data used for model building.

4. Folder - Images

• This folder contains all the graphs used in the Presentation Deck and Report.
• Codes to generate the graphs are recorded in the Notebook.

5. Folder - Models

• This folder contains the saved model developed for the project.
• The final model saved is used for production deployment as part of an application.