TRANSFORMING LAHORE'S REAL ESTATE MARKET WITH MACHINE LEARNING-DRIVEN PRICE PREDICTIONS

Abstract

This study aims to provide reliable information about the dynamics of the Lahore City, Pakistan real estate market and its response to economic changes, especially inflation. The study seeks to enhance forecasting methods for real estate prices in Lahore, Pakistan, to create a more transparent and efficient market. The research findings will guide decision-making processes and policy decisions aimed at promoting steady, long-term growth in Lahore's real estate market. The selling price of homes in Lahore, Pakistan was forecasted using various parametric regression models including the Extra Trees Repressor Model, XG-Boost Model, Random Forest Model, Gradient Boosting Model, Decision Tree Model, and Cat-Boost Regressor models. Data sourced from Zameen.com as of June 26, 2023, was utilized for this purpose. The dataset included 9539 homes located in prominent districts such as DHA Defense, Bahria Town, Johar Town, Park View City, Lake City, and Allama Iqbal Town. The average inflation rate in Pakistan's real estate market in 2024 was 24.76%. As graded by the R-square value and MSE, the Gradient Boosting (85%), and Extra Trees Regressor (85%) regression model emerged on top regarding the two metrics. In this dataset, it suffices to say that the Gradient Boosting and Extra Trees Regressor models are the way to go in predicting house prices.