COMPARATIVE ANALYSIS OF REGRESSION AND COLLABORATIVE FILTERING MODELS FOR RECOMMENDATION SYSTEMS: AN EMPIRICAL STUDY

Recommendation systems are very crucial for enhancing the utility of a given product and or service for a specific user in different fields. This paper focuses on the comparison of various filtering techniques in order to determine their effectiveness in identifying the preferences of the user. The paper also looks into basic methods like user-based collaborative filtering and item-based collaborative filtering, which uses the item's attributes. Also, the paper assesses the subsequent methods such as linear regression, ridge regression, Lasso regression, random forest regression, and XGBoost regression. From the performance evaluation metrics, the researchers reach RMSE and MAE to compare the effectiveness of the proposed methods and reveal their weaknesses. This paper aims to evaluate the performances of the above filtering approaches to gain an understanding of the extent to which these methods improve the recommendations' accuracy and contribute to the literature by providing recommendations on filtering models suitable for various recommendation tasks.