Predictive Modeling of Real Estate Prices Using Machine Learning Techniques
DOI:
https://doi.org/10.70917/ijcisim-2026-4497Keywords:
House Price Prediction, Machine Learning, Real Estate, Random Forest Regression, Linear Regression, Decision Tree Regression, K-Nearest Neighbors (KNN), Support Vector Regression (SVR)Abstract
Accurately estimating residential property prices is a challenging yet essential task in today’s changing market envi-ronment. Conventional manual valuation methods often result in inconsistencies; therefore, data-driven approaches have increased importance. This research focuses on developing a regression model using machine learning that can predict house prices and assist in informed decision-making within the real estate industry. The Ames, Iowa dataset from Kaggle contains 1,460 houses with 81 features that include lot size, neighborhood, construction year, and overall quality. Data pre processing involved handling missing values in numerical fields with median imputation and in categorical fields with mode imputation, followed by the application of one-hot encoding and standardization of features using the z-score.
The regression algorithms Random Forest, Linear Regression, Decision Tree, Support Vector Regression, and K-Nearest Neigh-bors were trained on an 80/20 train-test split and evaluated using the metrics R2, RMSE, and MAE. Among these, the Random Forest model performed best with R2 ≈ 0.891, RMSE ≈ 28,872, and MAE ≈ 17,614. On the other hand, the weakest performance was observed for Support Vector Regression.
The final Random Forest model was deployed as a web application using Streamlit, allowing users to input features and obtain instant predictions of property prices. Overall, the study showed that advanced regression methods and effective feature engineering significantly contribute to improving real estate price forecasts.