一种基于多正则化参数的矩阵分解推荐算法

Recommender algorithm based on matrix factorization with multiple regularization parameters

基于梯度下降矩阵分解模型的协同过滤推荐算法需要利用正则化技术对问题加以约束。损失函数中的正则化参数能够提高模型的预测精度,防止训练过拟合,并可以在二者间调节,使二者平衡。提出了一种多正则化参数的方法,根据用户的活跃度或者项目的流行度确定正则化参数的值,能在不同评分数量的用户或者项目上防止训练过拟合,同时可以得到更好的预测精度。实验结果验证了算法的正确性和有效性。

Regularization technique is needed for the collaborative filtering recommendation algorithm based on the gradientdescent matrix factorization to restrict the problems.Regularization parameter used in the loss function can control thetrade-off between prediction accuracy of the model and overfitting avoidance to the training.A method of multiple regularizationparameters is proposed.It can obtain the regularization parameters according to the user activity and item popularity,avoid the overfitting in the users and items with different numbers,and get better prediction accuracy.The experimentresults show that this method is correct and feasible.