Theoretically Accurate Regularization Technique for Matrix Factorization based Recommender Systems
This addresses a key bottleneck in recommender systems for users and developers, offering a more reliable regularization technique.
The paper tackles the problem of selecting regularization coefficients in matrix factorization for recommender systems, proving that common scalar-based approaches are invalid and proposing a theoretically accurate method that improves both accuracy and fairness metrics.
Regularization is a popular technique to solve the overfitting problem of machine learning algorithms. Most regularization technique relies on parameter selection of the regularization coefficient. Plug-in method and cross-validation approach are two most common parameter selection approaches for regression methods such as Ridge Regression, Lasso Regression and Kernel Regression. Matrix factorization based recommendation system also has heavy reliance on the regularization technique. Most people select a single scalar value to regularize the user feature vector and item feature vector independently or collectively. In this paper, we prove that such approach of selecting regularization coefficient is invalid, and we provide a theoretically accurate method that outperforms the most widely used approach in both accuracy and fairness metrics.