Session
Management Business and Economy
Description
The objective of the research was to perform a comparative evaluation of Linear Regression and Random Forest for predicting property prices in Prishtina. The analysis utilized a dataset comprising 1,512 property transactions, employing metrics such as Mean Squared Error (MSE), Coefficient of Determination (R²), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE) for evaluation. The study also included a variable importance assessment. After preprocessing and normalizing the data, both models were trained and evaluated. The findings indicated that the Random Forest model outperformed the Linear Regression model. The Variable Importance Test highlighted the distance from the central business district and the proximity to the main road leading to the central business district as key factors influencing housing prices in both models. This study emphasizes the importance of comparative analysis, meticulous data preparation, and the use of diverse evaluation metrics. It offers valuable insights for scientific research, practical application, and policy formulation, particularly in understanding efficient property price prediction mechanisms that could affect market equilibrium, real estate investment strategies, and housing policies. The uniqueness of this research lies in its methodological rigor and focus on the real estate market of Prishtina, a previously unexplored context, with potential implications for other emerging economies with dynamic real estate sectors such as Kosovo.
Keywords:
Machine Learning, Housing Prices Prediction, Linear Regression, Random Forest, Prishtina Real Estate Market, Comparative Analysis.
Proceedings Editor
Edmond Hajrizi
ISBN
978-9951-982-15-3
Location
UBT Kampus, Lipjan
Start Date
25-10-2024 9:00 AM
End Date
27-10-2024 6:00 PM
DOI
10.3107/ubt-ic.2024.27
Recommended Citation
Hoxha, Visar; Lecaj, Veli; Pallaska, Fuat; Dana, Hazer; Hoxha, Jehona; and Shala, Albana, "Comparative Analysis of Linear Regression and Random Forest for Property Price Prediction in Prishtina: Implications for Emerging Markets" (2024). UBT International Conference. 27.
https://knowledgecenter.ubt-uni.net/conference/2024UBTIC/MBE/27
Included in
Comparative Analysis of Linear Regression and Random Forest for Property Price Prediction in Prishtina: Implications for Emerging Markets
UBT Kampus, Lipjan
The objective of the research was to perform a comparative evaluation of Linear Regression and Random Forest for predicting property prices in Prishtina. The analysis utilized a dataset comprising 1,512 property transactions, employing metrics such as Mean Squared Error (MSE), Coefficient of Determination (R²), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE) for evaluation. The study also included a variable importance assessment. After preprocessing and normalizing the data, both models were trained and evaluated. The findings indicated that the Random Forest model outperformed the Linear Regression model. The Variable Importance Test highlighted the distance from the central business district and the proximity to the main road leading to the central business district as key factors influencing housing prices in both models. This study emphasizes the importance of comparative analysis, meticulous data preparation, and the use of diverse evaluation metrics. It offers valuable insights for scientific research, practical application, and policy formulation, particularly in understanding efficient property price prediction mechanisms that could affect market equilibrium, real estate investment strategies, and housing policies. The uniqueness of this research lies in its methodological rigor and focus on the real estate market of Prishtina, a previously unexplored context, with potential implications for other emerging economies with dynamic real estate sectors such as Kosovo.