Description of Invention
House price prediction is complex, influenced by factors like location, economy, and property features. Traditional methods often miss nonlinear patterns in real estate data. This study applies machine learning—using data from Kaggle, MyREI, and REHDA—to improve prediction accuracy. After preprocessing and EDA, models including Linear Regression, Random Forest, XGBoost, and SVR were tested. XGBoost performed best with an MAE of 0.12, RMSE of 0.18, and R² of 0.91. A web app using Streamlit or Flask was developed for users like buyers and agents. Despite data quality and local trend limitations, the approach shows strong potential for AI-driven real estate valuation.