Forecasting resale house prices using machine learning models

• Main Author: Krithika Jayaraman Karthikeyan
• Other Authors: Josephine Chong Leng Leng
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Computer and Information Science; Mathematical Sciences; Machine learning; Prediction
• Online Access: https://hdl.handle.net/10356/181191
• Institution: Nanyang Technological University

This study examines the effectiveness of various machine learning models while identifying key factors that influence the resale prices of HDB flats in Singapore, utilizing extensive datasets sourced from data.gov.sg and OneMap API. The analysis includes a broad range of features, such as lease commencement dates, flat characteristics, and proximity to essential amenities like MRT stations, schools, supermarkets, and hawker centers. Traditional machine learning models, including Random Forest, Gradient Boosting, and XGBoost, were applied to predict resale prices, using performance metrics such as R² and RMSE for evaluation. The results highlight the significant impact of demographic factors and location demograohics, providing valuable insights for home buyers and sellers. The findings were visualized in a Streamlit application, where the best-performing model was integrated to predict resale prices based on user inputs, offering an interactive tool for decision-making. This research deepens the understanding of housing market dynamics in Singapore and sets a foundation for future studies in property valuation.