A parallel matrix factorization based recommender by alternating stochastic gradient decent

A parallel matrix factorization based recommender by alternating stochastic gradient decent
复制标题

基于并行矩阵分解的交替随机梯度下降推荐器

DOI:
10.1016/j.engappai.2011.10.011
复制
发表时间:
2012-10
影响因子:
8
通讯作者:
朱庆生
朱庆生
中科院分区:
计算机科学2区
文献类型:
--
作者:
罗辛;刘慧君;夏云霓;朱庆生

文献摘要

参考文献

被引文献

相似文献

协同过滤(CF)可以通过矩阵分解(MF)实现,具有较高的预测精度和可扩展性。然而,目前大多数基于MF的并行编程器都是串行的,这使得它们无法分享并行编程技术的快速发展所带来的效率。针对基于正则化矩阵分解(Regularized Matrix Factorization,RMF)的CF推荐系统的并行化问题,首先对RMF的参数更新过程进行了理论分析,指出项目特征与用户特征之间的相互依赖性是阻碍该模型并行化的主要障碍。为了消除参数之间的相互依赖性,我们应用交替随机梯度求解器(ASGD)求解器来处理参数训练过程。在此基础上,我们随后提出了并行RMF(P-RMF)模型,其中的训练过程可以并行化,通过同时训练不同的用户/项目的功能。在两个大的真实的数据集上的实验表明,我们的P-RMF模型可以提供一个更快的解决CF问题相比,原来的RMF和另一个并行MF推荐。
Collaborative Filtering (CF) can be achieved by Matrix Factorization (MF) with high prediction accuracy and scalability. Most of the current MF based recommenders, however, are serial, which prevent them sharing the efficiency brought by the rapid progress in parallel programming techniques. Aiming at parallelizing the CF recommender based on Regularized Matrix Factorization (RMF), we first carry out the theoretical analysis on the parameter updating process of RMF, whereby we can figure out that the main obstacle preventing the model from parallelism is the inter-dependence between item and user features. To remove the inter-dependence among parameters, we apply the Alternating Stochastic Gradient Solver (ASGD) solver to deal with the parameter training process. On this basis, we subsequently propose the parallel RMF (P-RMF) model, of which the training process can be parallelized through simultaneously training different user/item features. Experiments on two large, real datasets illustrate that our P-RMF model can provide a faster solution to CF problem when compared to the original RMF and another parallel MF based recommender.
DOI: --
发表时间: 2010-12
期刊: --
影响因子: --
作者:
Gilles Louppe;P. Geurts
通讯作者: Gilles Louppe;P. Geurts
DOI: 10.21236/ada439541
发表时间: 2000-07
期刊: Brain Stimulation
影响因子: 7.7
作者:
B. Sarwar;G. Karypis;J. Konstan;J. Riedl
通讯作者: B. Sarwar;G. Karypis;J. Konstan;J. Riedl
DOI: 10.1002/9781119183952.ch14
发表时间: 2018-01
期刊: --
影响因子: --
作者:
M. Krochmal;H. Husi
通讯作者: M. Krochmal;H. Husi
DOI: --
发表时间: 2006-04
期刊: --
影响因子: --
作者:
G. Gorrell
通讯作者: G. Gorrell
DOI: --
发表时间: 2007
期刊: --
影响因子: --
作者:
Miklós Kurucz;A. Benczúr;Károly Csalogány
通讯作者: Miklós Kurucz;A. Benczúr;Károly Csalogány