Link to Medium article (Part 1 - Web Scraping and EDA): https://medium.com/swlh/web-scrapping-and-data-analysis-of-condominium-rental-market-in-singapore-da5265c71d19
Link to Medium article (Part 2 - Prediction with Ensemble Regressors and MLR): https://medium.com/datadriveninvestor/using-ensemble-regressors-to-predict-condo-rental-prices-47eb7c3d5cd9
- There are plenty of analytical write-ups on predicting prices of public housing sales (especially the Boston public housing dataset). However, I have not really come across any projects describing the private property market.
- In this project, I worked on exploring and predicting the rental prices in the private property market (specifically condominiums in Singapore)
- There are 3 parts to this project, namely:
- Data acquisition (use of web scrapping) to curate data related to condominium rental
- Data pre-processing and exploratory data analysis (EDA) to analyze and identify trends in the condominium rental market
- Prediction of rental price with 5 different ML models i.e. Linear Regression, Random Forest Regressor, XGBoost Regressor, LightGBM Regressor, and CatBoost Regressor
- This project is an end-to-end endeavour, starting from the beginning where web scrapping was utilized to acquire data, all the way to running machine learning models to predict rental prices.
- There is a lack of projects analyzing the private property market, in particular the market of condominium rentals in Singapore. This project serves to address this gap.
- Insights of the condominium rental market are visualized and explained, and detailed codes of all the steps are included in the notebooks. In addition, I also demonstrated how to perform checks of the key assumptions of linear regression models, and described the general concepts of ensemble learning.