What Drives House Prices? A Linear Regression Approach to Size, Condition, and Features

This research examines the key factors influencing house prices, focusing on how size, condition, and structural features contribute to property valuation. Using a dataset from Washington State, USA, covering the year ۲۰۱۴ with over ۴,۶۰۰ entries, a multivariate analysis was conducted with a Linear Regression model to assess the relationships between crucial features such as square footage, number of bedrooms, bathrooms, floors, and additional structural elements like garage presence and yard size. The analysis revealed that square footage and bathrooms exhibit the strongest positive correlations with house prices (both with correlation values of ۰.۷۶, statistically significant at p < ۰.۰۵), indicating their substantial impact on property valuation. In contrast, factors like condition and view demonstrated weaker correlations, suggesting a more limited influence. The Linear Regression model explained ۷۵% of the variation in house prices (R۲ = ۰.۷۵), with validation conducted using a holdout test set to ensure generalizability. While the model effectively highlights key price determinants, its limitations in handling non-linear relationships and sensitivity to outliers were addressed through data transformation and outlier removal. Compared to prior studies, this research reinforces established findings on square footage and bathrooms while providing new insights into the comparatively lower impact of property condition. Future work could explore advanced predictive models, such as non-linear regression and machine learning techniques, to better capture complex relationships and improve forecasting accuracy. These findings offer valuable insights for buyers, sellers, and industry professionals, emphasizing the importance of a data-driven approach to understanding house price dynamics.

کلیدواژه ها:

House Price Prediction ، Linear Regression ، multivariate analysis ، Property Features ، Market Valuation