|
|||||
|
|
| 基于KNN-SVM的混合协同过滤推荐算法 | |
| 引用本文： | 吕成戍,王维国,丁永健.基于KNN-SVM的混合协同过滤推荐算法[J].计算机应用研究,2012,29(5):1707-1709. |
| 作者姓名： | 吕成戍 王维国 丁永健 |
| 作者单位： | 1. 东北财经大学管理科学与工程学院,辽宁大连,116025 2. 东北财经大学数学与数量经济学院,辽宁大连,116025 3. 大连理工大学管理与经济学部,辽宁大连,116024 |
| 基金项目： | 辽宁省社会科学规划基金资助项目（L10BJL035）；中央高校专项科研基金资助项目（DUT10RW302） |
| 摘 要： | 数据稀疏性问题对协同过滤推荐系统的推荐精度有很大影响,为此,融合缺失数据平衡方法,提出了一个基于KNN-SVM的混合协同过滤推荐算法。利用K-最近邻法对训练集中的缺失数据进行填补,然后通过支持向量机交叉验证进行分类,综合两者优点,从而克服数据质量对推荐算法的影响。在标杆数据集上进行了仿真实验,数值结果证明了方法的有效性。 |
| 关 键 词： | 数据稀疏性 支持向量机 K-最近邻 协同过滤 |
Hybrid collaborative filtering algorithm based on KNN-SVM |
|
| LV Cheng-shu,WANG Wei-guo,DING Yong-jian.Hybrid collaborative filtering algorithm based on KNN-SVM[J].Application Research of Computers,2012,29(5):1707-1709. | |
| Authors: | LV Cheng-shu WANG Wei-guo DING Yong-jian |
| Affiliation: | 1. a. School of Management Science & Engineering, b. School of Mathematics & Quantitative Economics, Dongbei University of Finance & Economics, Dalian Liaoning 116025, China; 2. Faculty of Management & Economics, Dalian University of Technology, Dalian Liaoning 116024, China |
| Abstract: | The problem of data sparseness has great influence on collaborative filtering recommendation system's accuracy, balance for this missing data fusion method, this paper proposed a hybrid collaborative filtering algorithms based on KNN-SVM. K-nearest neighbor method used the training set to fill the missing data, and then cross-validated by SVM classification. Comprehend advantages both KNN and SVM in order to overcome impact of data quality on the recommended algorithm. The proposed approach was applied to benchmark problems, and the simulation results show it is valid. |
| Keywords: | data sparsity support vector machine(SVM) K-nearest neighbor collaborative filtering |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |
| 点击此处可从《计算机应用研究》浏览原始摘要信息 | |
| 点击此处可从《计算机应用研究》下载全文 | |