Efficient Preference Clustering via Random Fourier Features
Approximations based on random Fourier features have recently emerged as an efficient and elegant method for designing large-scale machine learning tasks. Unlike approaches using the Nyström method, which randomly samples the training examples, we make use of random Fourier features, whose basis fun...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Tsinghua University Press
2019-09-01
|
| Series: | Big Data Mining and Analytics |
| Subjects: | |
| Online Access: | https://www.sciopen.com/article/10.26599/BDMA.2019.9020003 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832572847836364800 |
|---|---|
| author | Jingshu Liu Li Wang Jinglei Liu |
| author_facet | Jingshu Liu Li Wang Jinglei Liu |
| author_sort | Jingshu Liu |
| collection | DOAJ |
| description | Approximations based on random Fourier features have recently emerged as an efficient and elegant method for designing large-scale machine learning tasks. Unlike approaches using the Nyström method, which randomly samples the training examples, we make use of random Fourier features, whose basis functions (i.e., cosine and sine ) are sampled from a distribution independent from the training sample set, to cluster preference data which appears extensively in recommender systems. Firstly, we propose a two-stage preference clustering framework. In this framework, we make use of random Fourier features to map the preference matrix into the feature matrix, soon afterwards, utilize the traditional k-means approach to cluster preference data in the transformed feature space. Compared with traditional preference clustering, our method solves the problem of insufficient memory and greatly improves the efficiency of the operation. Experiments on movie data sets containing 100 000 ratings, show that the proposed method is more effective in clustering accuracy than the Nyström and k-means, while also achieving better performance than these clustering approaches. |
| format | Article |
| id | doaj-art-f1f6f716b5c34b6a9b22bc026fcad841 |
| institution | Kabale University |
| issn | 2096-0654 |
| language | English |
| publishDate | 2019-09-01 |
| publisher | Tsinghua University Press |
| record_format | Article |
| series | Big Data Mining and Analytics |
| spelling | doaj-art-f1f6f716b5c34b6a9b22bc026fcad8412025-02-02T06:50:33ZengTsinghua University PressBig Data Mining and Analytics2096-06542019-09-012319520410.26599/BDMA.2019.9020003Efficient Preference Clustering via Random Fourier FeaturesJingshu Liu0Li Wang1Jinglei Liu2<institution content-type="dept">College of Data Science</institution>, <institution>Taiyuan University of Technology</institution>, <city>Jinzhong</city> <postal-code>030600</postal-code>, <country>China</country>.<institution content-type="dept">College of Data Science</institution>, <institution>Taiyuan University of Technology</institution>, <city>Jinzhong</city> <postal-code>030600</postal-code>, <country>China</country>.<institution content-type="dept">School of Computer and Control Engineering</institution>, <institution>Yantai University</institution>, <city>Yantai</city> <postal-code>264005</postal-code>, <country>China</country>.Approximations based on random Fourier features have recently emerged as an efficient and elegant method for designing large-scale machine learning tasks. Unlike approaches using the Nyström method, which randomly samples the training examples, we make use of random Fourier features, whose basis functions (i.e., cosine and sine ) are sampled from a distribution independent from the training sample set, to cluster preference data which appears extensively in recommender systems. Firstly, we propose a two-stage preference clustering framework. In this framework, we make use of random Fourier features to map the preference matrix into the feature matrix, soon afterwards, utilize the traditional k-means approach to cluster preference data in the transformed feature space. Compared with traditional preference clustering, our method solves the problem of insufficient memory and greatly improves the efficiency of the operation. Experiments on movie data sets containing 100 000 ratings, show that the proposed method is more effective in clustering accuracy than the Nyström and k-means, while also achieving better performance than these clustering approaches.https://www.sciopen.com/article/10.26599/BDMA.2019.9020003random fourier featuresmatrix decompositionsimilarity matrix |
| spellingShingle | Jingshu Liu Li Wang Jinglei Liu Efficient Preference Clustering via Random Fourier Features Big Data Mining and Analytics random fourier features matrix decomposition similarity matrix |
| title | Efficient Preference Clustering via Random Fourier Features |
| title_full | Efficient Preference Clustering via Random Fourier Features |
| title_fullStr | Efficient Preference Clustering via Random Fourier Features |
| title_full_unstemmed | Efficient Preference Clustering via Random Fourier Features |
| title_short | Efficient Preference Clustering via Random Fourier Features |
| title_sort | efficient preference clustering via random fourier features |
| topic | random fourier features matrix decomposition similarity matrix |
| url | https://www.sciopen.com/article/10.26599/BDMA.2019.9020003 |
| work_keys_str_mv | AT jingshuliu efficientpreferenceclusteringviarandomfourierfeatures AT liwang efficientpreferenceclusteringviarandomfourierfeatures AT jingleiliu efficientpreferenceclusteringviarandomfourierfeatures |