Benchmarking Predictive Performance Of Machine Learning Approaches For Accurate Prediction Of Boston House Prices: An In-Depth Analysis

Abstract

The global market for real estate is quite large. Over the last several decades, there has been a notable increase in this domain. Consumers and other decision-makers may make better choices if an accurate forecast is made. Nonetheless, it remains difficult to create a model that can accurately forecast home values in such situations. This study paper suggests a house price prediction model based on ML that can better guess house prices. A mix of data pre-processing methods and ML algorithms are used at the same time in the suggested model. Using the Real Estate Dataset to test how well the proposed model works, the results show that it is much better than the current methods. The research highlights how important it is to handle null values and remove outliers from data in order to get better results. This study has attempted to implement various machine learning algorithms like XGB, GB, and LGBM algorithms. The models' performance is evaluated by employing metrics like MSE, MAE, Mean, RMSE, and standard deviation on the test dataset. From the experimental outcomes, the XGB model achieved RMSE= 2.45, MSE= 6.03, MAE=1.32, Mean=21.04, and standard deviation = 3.77. Based on these findings, advocate employing the DL technique for evaluating property values in future study.