COMPARATIVE ANALYSIS OF MACHINE LEARNING ALGORITHMS FOR HOUSE PRICE PREDICTION

This study developed a reliable and accurate house price prediction system to support decision making in the real estate sector, where proper valuation plays a vital role for buyers, sellers, investors, and policymakers. The objective was to apply machine learning techniques to design a data driven framework capable of improving prediction accuracy, transparency, and efficiency in housing valuation. The research was conducted using a Kaggle dataset containing Lagos housing data and employed four supervised learning models: Linear Regression, Decision Tree, Random Forest, and XGBoost. Prior to model training, the dataset underwent preprocessing steps such as handling missing values, encoding categorical variables, normalisation, and outlier detection. The theoretical framework was based on predictive modelling using machine learning, and model performance was evaluated with Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and R² metrics. The Random Forest model achieved the highest performance with an R² of 0.66, outperforming the other models and demonstrating stronger predictive capability and robustness. The system was implemented in Python and deployed via a Streamlit interface, which enables users to input housing features and obtain real time price predictions accompanied by visual analytics. Based on these findings, the study establishes that ensemble models are effective in capturing complex and non linear interactions within housing data, thereby improving predictive accuracy. The study demonstrates the potential of machine learning to transform real estate valuation by offering scalable and objective solutions, and it recommends the integration of predictive analytics tools into real estate markets to enhance decision support.