<![CDATA[Significant differences in home prices, even on properties with similar building sizes and locations, pose a major challenge in accurately determining property valuations. The discrepancy between the actual market price and the estimated value makes it difficult for potential buyers, sellers, and developers to make the right decision. To overcome these problems, this study applied the Multiple Linear Regression (MLR) algorithm in the Decision Support System (DSS) to estimate house prices based on the location and area of the building. The dataset used consists of 545 housing data points with variables such as house prices, locations, and building areas. The research stages include data collection, pre-processing (data cleaning and normalization), model development using MLR, and model performance evaluation. The evaluation was carried out using the division of trained data and test data with an 80:20 ratio, so that the model was tested using data that was not previously trained. The results showed that the model produced a Mean Absolute Error (MAE) of 1,474,748.13, a Root Mean Squared Error (RMSE) of 1,917,103.70, and a coefficient of determination (R²) of 0.273. A relatively low R² value indicates that the location and area variables of the building are not sufficient to explain the overall variation in house prices, so the addition of other variables—such as the number of rooms, facilities, and environmental conditions—is needed to improve the accuracy of the prediction and produce a more representative price estimate.]]>
Copyrights © 2025
