Genetic Algorithm Based Intelligent System for Estate Value Estimation

Abstract

The prices of plots are increasing day by day, so prediction of plot prices has been a challenging task in recent times. In this modern age,  people want an intelligent prediction system of plot price which can predict plot price accurately. For this purpose, we need a proper model which estimates the plot price. People are worried because of the increasing plots rates, and they want to buy a  plot with their budget. In our proposed methodology, we have used a genetic algorithm based on optimistic feature selection to improve results. As we know, feature selection plays a vital role in computational models of machine learning.  However,  the predicted price should be estimated in the current scenario. This research presents the usage of linear, lasso, ridge, and decision tree regression techniques to predict the plot prices of eight different housing colonies of Multan, Pakistan. Dataset consists of 13 features that are chosen according to the general requirements and commonly popular demands. The system will be helpful for people to buy a plot within their limited budget. In this research, a genetic algorithm is also applied for feature selection. The best features selected by the G.A. have shown better performance in the prediction by classifiers. Machine Learning classifiers have shown better performance after best feature selection.