A Gradient Boosting Method for Effective Prediction of Housing Prices in Complex Real Estate Systems

Abstract

Analyzing real estate market changes by different parties and agencies that have a significant effect on real estate health and trends. In complex real estate systems, the prediction of housing prices plays an important role in mitigating the impacts of property valuation and economic growth. Several works have proposed the use of various machine learning models for predicting housing prices of real estate markets. However, developing an effective machine learning models to predict the housing prices is still a challenge and needs to be investigated. Therefore, this paper proposes an optimized model based on the gradient boosting (GB) method for improving the prediction of housing prices in complex real estate systems. To evaluate the proposed method, a set of experiments is conducted on a public real estate dataset. The experimental results show that the optimized GB (OGB) method can be used effectively for housing price prediction of real estate and achieves 0.01167 of the root mean square error; the lowest result compared to the other baseline machine learning models.