The Use of Modified Extreme Boosting in Machine Learning for the Prediction of Home Prices

Abstract

Machine learning's impact on commonplace spoken commands and predictions has grown in recent years. In its place, it offers an improved customer service system and a safer automated driving experience. The evidence suggests that ML is a promising technology with the potential to revolutionize many different markets. The home Price Index is a common tool used to measure changes in home values (HPI). A person's home price cannot be reliably predicted using only the HPI because of the strong relationship between property prices and other factors like population, area, and location. While some research has been able to accurately forecast home values using traditional machine learning methods, these studies almost never compare and contrast various models and completely exclude the more intricate but less well-known ones. The adaptive and probabilistic model selection procedure of Modified Extreme Gradient Boosting led us to suggest it as our model for this investigation. This procedure includes creating features, training and optimizing hyperparameters, interpreting models, and finally, selecting and evaluating models. House price indices, which are often used to bolster housing policy efforts and provide cost estimates. Using machine learning techniques, this research builds models to predict future changes in house values. Location, square footage, home price, and modified extreme gradient boosting are all relevant terms.