摘要
针对传统的基于最近邻协同过滤推荐算法中计算相似度存在的缺陷,提出了一种基于二阶段相似度学习的协同过滤推荐算法,该算法旨在通过较少的迭代计算改善推荐算法性能。它以既约梯度法迭代寻优为主、最近邻算法为辅,通过邻居的海选和精选,最终提高了相似度的计算精度,改善了误差性能。实验表明,在一定条件下该算法不仅在误差性能上优于传统的推荐算法,而且其算法收敛速度快,可实现相似度参数动态调整和分布式计算。
In order to improve the accuracy of similarity calculation and recommendation performance in the traditional collabo- rative filtering recommender system, this paper proposed a collaborative filtering recommendation algorithm based on two stages of similarity learning. The algorithm took advantage of the nearest neighbor algorithm on the first stage to get candidate neigh- bors and used the reduced gradient method on the second stage to learn similarity. Eventually, the algorithm achieved a higher accuracy of similarity, The experimental results show that the proposed algorithm, on some conditions, not only outperforms the traditional method in terms of the error performance, but also has a fast convergence speed, which can make dynamic simi- larity adjustment and distributed calculation possible.
出处
《计算机应用研究》
CSCD
北大核心
2013年第3期715-719,共5页
Application Research of Computers
基金
国家"863"计划资助项目(2011AA040605)
关键词
二阶段
相似度学习
协同过滤
既约梯度法
K-最近邻算法
two stages
similarity learning
collaborative filtering
reduced gradient method
K-nearest neighbor（ K-NN ）
