Predicting New York Housing Prices: A Comparative Analysis of Machine Learning Models

Jie Yu

2024

Abstract

Accurate prediction of housing prices in New York City is crucial for investors, policymakers, and consumers navigating one of the most volatile housing markets. This study explores various machine learning methods to forecast housing prices in New York City. The predictive power of Linear Regression (LR), Support Vector Regression (SVR), Random Forest (RF), and XGBoost (XGB) was examined using a comprehensive dataset with diverse housing attributes. Our results show that LR and SVR provided less accurate predictions, with LR achieving an RMSE of 4,091,594, a MAPE of 1.2991, and an adjusted R-squared of 0.2642, while SVR had an RMSE of 4,967,168, a MAPE of 0.7753, and an adjusted R-squared of -0.0844. In contrast, ensemble methods, namely RF and XGB, demonstrated superior performance on all accounts. RF achieved an RMSE of 2,145,123, a MAPE of 0.3086, and an adjusted R-squared of 0.7978, while XGB had an RMSE of 2,483,884, a MAPE of 0.4163, and an adjusted R-squared of 0.7288. These results conclude that ensemble methods, which can handle complex datasets with higher dimensionality and noise, are more adept at predicting housing prices in varied markets such as New York City. The findings have implications for stakeholders in the real estate industry seeking to leverage machine learning for investment and policy-making decisions.

Paper Citation

in Harvard Style

Yu J. (2024). Predicting New York Housing Prices: A Comparative Analysis of Machine Learning Models. In Proceedings of the 1st International Conference on Innovations in Applied Mathematics, Physics and Astronomy - Volume 1: IAMPA; ISBN 978-989-758-722-1, SciTePress, pages 102-109. DOI: 10.5220/0012999000004601

in Bibtex Style

@conference{iampa24,
author={Jie Yu},
title={Predicting New York Housing Prices: A Comparative Analysis of Machine Learning Models},
booktitle={Proceedings of the 1st International Conference on Innovations in Applied Mathematics, Physics and Astronomy - Volume 1: IAMPA},
year={2024},
pages={102-109},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012999000004601},
isbn={978-989-758-722-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Innovations in Applied Mathematics, Physics and Astronomy - Volume 1: IAMPA
TI - Predicting New York Housing Prices: A Comparative Analysis of Machine Learning Models
SN - 978-989-758-722-1
AU - Yu J.
PY - 2024
SP - 102
EP - 109
DO - 10.5220/0012999000004601
PB - SciTePress