Machine Learning Based House Price Prediction Using Modified Extreme Boosting

Abstract

In recent years, machine learning has become increasingly important in everyday voice commands and predictions. Instead, it provides a safer auto system and better customer assistance. As a result of all that has been demonstrated, ML is a technology that is becoming more and more popular in a range of industries. To gauge changes in house values, the House Price Index is frequently employed (HPI). Due to the substantial correlation that exists between property prices and other variables, such as location, region, and population, the HPI on its own is not sufficient to accurately forecast a person's house price. Some studies have successfully predicted house prices using conventional machine learning techniques, but they seldom evaluate the efficacy of different models and ignore the more complicated but less well-known models. We proposed Modified Extreme Gradient Boosting as our model in this study due to its adaptive and probabilistic model selection process. Feature engineering, hyperparameter training and optimization, model interpretation, and model selection and evaluation are all steps in the process. Home price indices, which are frequently used to support real estate policy initiatives and estimate housing costs. In this project, models for forecasting changes in home prices are developed using machine learning methods.