1. NCCU Academic Hub
  2. Publications
  3. College of Commerce
  4. Theses
  5. Item

Publications-Theses

Article View/Open

Publication Export

Google ScholarTM

NCCU Library

Citation Infomation

Related Publications in TAIR

題名 推薦系統資料插補改良法-電影推薦系統應用
Improving recommendations through data imputation-with application for movie recommendation
作者 楊智博
Yang, Chih Po
貢獻者 翁久幸
Weng, Chiu Hsing
楊智博
Yang, Chih Po
關鍵詞 推薦系統
矩陣分解
隨機梯度下降
奇異值分解
Recommender systems
Matrix Factorization
Stochastic Gradient Descent
Alternating Least Squares
Singular Value Decomposition
日期 2015
上傳時間 1-Oct-2015 14:12:21 (UTC+8)
摘要 現今許多網路商店或電子商務將產品銷售給消費者的過程中，皆使用推薦系統的幫助來提高銷售量。如亞馬遜公司(Amazon)、Netflix，深入了解顧客的使用習慣，建構專屬的推薦系統並進行個性化的推薦商品給每一位顧客。
推薦系統應用的技術分為協同過濾和內容過濾兩大類，本研究旨在探討協同過濾推薦系統中潛在因子模型方法，利用矩陣分解法找出評分矩陣。在Koren等人(2009)中，將矩陣分解法的演算法大致分為兩種，隨機梯度下降法(Stochastic gradient descent)與交替最小平方法(Alternating least squares)。本研究主要研究目的有三項，一為比較交替最小平方法與隨機梯度下降法的預測能力，二為兩種矩陣分解演算法在加入偏誤項後的表現，三為先完成交替最小平方法與隨機梯度下降法，以其預測值對原始資料之遺失值進行資料插補，再利用奇異值分解法對完整資料做矩陣分解，觀察其前後方法的差異。
研究結果顯示，隨機梯度下降法所需的運算時間比交替最小平方法所需的運算時間少。另外，完成兩種矩陣分解演算法後，將預測值插補遺失值，進行奇異值分解的結果也顯示預測能力有提升。
Recommender system has been largely used by Internet companies such Amazon and Netflix to make recommendations for Internet users. Techniques for recommender systems can be divided into content filtering approach and collaborative filtering approach. Matrix factorization is a popular method for collaborative filtering approach. It minimizes the object function through stochastic gradient descent and alternating least squares.
This thesis has three goals. First, we compare the alternating least squares method and stochastic gradient descent method. Secondly, we compare the performance of matrix factorization method with and without the bias term. Thirdly, we combine singular value decomposition and matrix factorization.
As expected, we found the stochastic gradient descent takes less time than the alternating least squares method, and the the matrix factorization method with bias term gives more accurate prediction. We also found that combining singular value decomposition with matrix factorization can improve the predictive accuracy.
參考文獻 1. Resnick, Paul, et al. "GroupLens: an open architecture for collaborative filtering of netnews." Proceedings of the 1994 ACM conference on Computer supported cooperative work. ACM,1994.

2. Konstan, Joseph A., et al. "GroupLens: applying collaborative filtering to Usenet news." Communications of the ACM 40.3,1997,pp. 77-87.

3. Sarwar, Badrul, et al. "Item-based collaborative filtering recommendation algorithms." Proceedings of the 10th international conference on World Wide Web. ACM, 2001.

4. Linden, Greg, Brent Smith, and Jeremy York. "Amazon. com recommendations: Item-to-item collaborative filtering." Internet Computing, IEEE 7.1,2003,pp. 76-80.

5. Paterek, Arkadiusz. "Improving regularized singular value decomposition for collaborative filtering." Proceedings of KDD cup and workshop. Vol. 2007. 2007.

6. Takács, Gábor, et al. "Major components of the gravity recommendation system." ACM SIGKDD Explorations Newsletter 9.2,2007,pp. 80-83.

7. Takacs, Gabor, et al. "On the gravity recommendation system." Proceedings of KDD cup and workshop. Vol. 2007. 2007.


8. Ma, Chih-Chao. "A Guide to Singular Value Decomposition for Collaborative Filtering.",2008.

9. Koren, Yehuda. "The bellkor solution to the netflix grand prize." Netflix prize documentation 81,2009.

10. Koren, Yehuda, Robert Bell, and Chris Volinsky. "Matrix factorization techniques for recommender systems." Computer 8,2009,pp. 30-37.

11. Gong, Songjie. "A collaborative filtering recommendation algorithm based on user clustering and item clustering." Journal of Software 5.7,2010,pp. 745-752.

12. Koren, Yehuda. "Collaborative filtering with temporal dynamics."Communications of the ACM 53.4,2010,pp. 89-97.

13. Bottou, Léon. "Large-scale machine learning with stochastic gradient descent."Proceedings of COMPSTAT`2010. Physica-Verlag HD, 2010,pp. 177-186.


14. 吳金龍,Netflix Prize 中的協同過濾算法,北京大學數學科學學院博士論文,2010

15. Ekstrand, Michael D., John T. Riedl, and Joseph A. Konstan. "Collaborative filtering recommender systems." Foundations and Trends in Human-Computer Interaction 4.2,2011,pp. 81-173.

16. 張孫浩,網路評比資料之統計分析,國立政治大學統計學系碩士論文,2011

17. Bottou, Léon. "Stochastic gradient descent tricks." Neural Networks: Tricks of the Trade. Springer Berlin Heidelberg, 2012,pp. 421-436.

18. Zhuang, Yong, et al. "A fast parallel sgd for matrix factorization in shared memory systems." Proceedings of the 7th ACM conference on Recommender systems. ACM, 2013.

19. 張良卉,矩陣分解法對網路評比資料分析之探討,國立政治大學統計學系碩士論文,2013

20. Fang, Xiaowen. A Study of Recommender Systems with Applications. Diss. UNIVERSITY OF MINNESOTA,2014.

21. GroupLens Research. Retrieved JUN,2015,from http://www.grouplens.org
描述 碩士
國立政治大學
統計研究所
102354020
資料來源 http://thesis.lib.nccu.edu.tw/record/#G0102354020
資料類型 thesis
dc.contributor.advisor 翁久幸zh_TW
dc.contributor.advisor Weng, Chiu Hsingen_US
dc.contributor.author (Authors) 楊智博zh_TW
dc.contributor.author (Authors) Yang, Chih Poen_US
dc.creator (作者) 楊智博zh_TW
dc.creator (作者) Yang, Chih Poen_US
dc.date (日期) 2015en_US
dc.date.accessioned 1-Oct-2015 14:12:21 (UTC+8)-
dc.date.available 1-Oct-2015 14:12:21 (UTC+8)-
dc.date.issued (上傳時間) 1-Oct-2015 14:12:21 (UTC+8)-
dc.identifier (Other Identifiers) G0102354020en_US
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/78727-
dc.description (描述) 碩士zh_TW
dc.description (描述) 國立政治大學zh_TW
dc.description (描述) 統計研究所zh_TW
dc.description (描述) 102354020zh_TW
dc.description.abstract (摘要) 現今許多網路商店或電子商務將產品銷售給消費者的過程中，皆使用推薦系統的幫助來提高銷售量。如亞馬遜公司(Amazon)、Netflix，深入了解顧客的使用習慣，建構專屬的推薦系統並進行個性化的推薦商品給每一位顧客。
推薦系統應用的技術分為協同過濾和內容過濾兩大類，本研究旨在探討協同過濾推薦系統中潛在因子模型方法，利用矩陣分解法找出評分矩陣。在Koren等人(2009)中，將矩陣分解法的演算法大致分為兩種，隨機梯度下降法(Stochastic gradient descent)與交替最小平方法(Alternating least squares)。本研究主要研究目的有三項，一為比較交替最小平方法與隨機梯度下降法的預測能力，二為兩種矩陣分解演算法在加入偏誤項後的表現，三為先完成交替最小平方法與隨機梯度下降法，以其預測值對原始資料之遺失值進行資料插補，再利用奇異值分解法對完整資料做矩陣分解，觀察其前後方法的差異。
研究結果顯示，隨機梯度下降法所需的運算時間比交替最小平方法所需的運算時間少。另外，完成兩種矩陣分解演算法後，將預測值插補遺失值，進行奇異值分解的結果也顯示預測能力有提升。		zh_TW
dc.description.abstract (摘要) Recommender system has been largely used by Internet companies such Amazon and Netflix to make recommendations for Internet users. Techniques for recommender systems can be divided into content filtering approach and collaborative filtering approach. Matrix factorization is a popular method for collaborative filtering approach. It minimizes the object function through stochastic gradient descent and alternating least squares.
This thesis has three goals. First, we compare the alternating least squares method and stochastic gradient descent method. Secondly, we compare the performance of matrix factorization method with and without the bias term. Thirdly, we combine singular value decomposition and matrix factorization.
As expected, we found the stochastic gradient descent takes less time than the alternating least squares method, and the the matrix factorization method with bias term gives more accurate prediction. We also found that combining singular value decomposition with matrix factorization can improve the predictive accuracy.		en_US
dc.description.tableofcontents 第一章 緒論 1
1.1研究背景 1
1.2推薦系統介紹 2
1.2.1 內容導向過濾(Content-Based Filtering) 2
1.2.3 協同過濾(Collaborative Filtering) 3
1.2.4 混合過濾(Hybrid Filtering) 4
1.3 本論文研究目的 4
第二章 文獻回顧 5
2.1 鄰域法 5
2.1.1 鄰域法-基於使用者(User-Based) 5
2.1.2 鄰域法-基於項目(Item-Based) 7
2.2 潛在因子模型 9
第三章 研究方法 11
3.1 矩陣分解法 11
3.1.1 奇異值分解法 11
3.1.2 潛在因子模型 13
3.2 隨機梯度下降法 17
3.2.1 基礎概念 17
3.2.2 實例說明 18
3.3 預測指標與收斂條件 22
第四章 實證研究 24
4.1 無偏誤項的模型比較 24
4.2 有偏誤項的模型比較 26
4.3 不同資料量的分析 28
4.4 二階段矩陣分解：插補矩陣與奇異值分解 30
4.4.1 基本想法與改進方法步驟 30
4.4.2 結果 33
第五章 結論與建議 35
參考文獻及相關目錄 37		zh_TW
dc.format.extent 1043828 bytes-
dc.format.mimetype application/pdf-
dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0102354020en_US
dc.subject (關鍵詞) 推薦系統zh_TW
dc.subject (關鍵詞) 矩陣分解zh_TW
dc.subject (關鍵詞) 隨機梯度下降zh_TW
dc.subject (關鍵詞) 奇異值分解zh_TW
dc.subject (關鍵詞) Recommender systemsen_US
dc.subject (關鍵詞) Matrix Factorizationen_US
dc.subject (關鍵詞) Stochastic Gradient Descenten_US
dc.subject (關鍵詞) Alternating Least Squaresen_US
dc.subject (關鍵詞) Singular Value Decompositionen_US
dc.title (題名) 推薦系統資料插補改良法-電影推薦系統應用zh_TW
dc.title (題名) Improving recommendations through data imputation-with application for movie recommendationen_US
dc.type (資料類型) thesisen
dc.relation.reference (參考文獻) 1. Resnick, Paul, et al. "GroupLens: an open architecture for collaborative filtering of netnews." Proceedings of the 1994 ACM conference on Computer supported cooperative work. ACM,1994.

2. Konstan, Joseph A., et al. "GroupLens: applying collaborative filtering to Usenet news." Communications of the ACM 40.3,1997,pp. 77-87.

3. Sarwar, Badrul, et al. "Item-based collaborative filtering recommendation algorithms." Proceedings of the 10th international conference on World Wide Web. ACM, 2001.

4. Linden, Greg, Brent Smith, and Jeremy York. "Amazon. com recommendations: Item-to-item collaborative filtering." Internet Computing, IEEE 7.1,2003,pp. 76-80.

5. Paterek, Arkadiusz. "Improving regularized singular value decomposition for collaborative filtering." Proceedings of KDD cup and workshop. Vol. 2007. 2007.

6. Takács, Gábor, et al. "Major components of the gravity recommendation system." ACM SIGKDD Explorations Newsletter 9.2,2007,pp. 80-83.

7. Takacs, Gabor, et al. "On the gravity recommendation system." Proceedings of KDD cup and workshop. Vol. 2007. 2007.


8. Ma, Chih-Chao. "A Guide to Singular Value Decomposition for Collaborative Filtering.",2008.

9. Koren, Yehuda. "The bellkor solution to the netflix grand prize." Netflix prize documentation 81,2009.

10. Koren, Yehuda, Robert Bell, and Chris Volinsky. "Matrix factorization techniques for recommender systems." Computer 8,2009,pp. 30-37.

11. Gong, Songjie. "A collaborative filtering recommendation algorithm based on user clustering and item clustering." Journal of Software 5.7,2010,pp. 745-752.

12. Koren, Yehuda. "Collaborative filtering with temporal dynamics."Communications of the ACM 53.4,2010,pp. 89-97.

13. Bottou, Léon. "Large-scale machine learning with stochastic gradient descent."Proceedings of COMPSTAT`2010. Physica-Verlag HD, 2010,pp. 177-186.


14. 吳金龍,Netflix Prize 中的協同過濾算法,北京大學數學科學學院博士論文,2010

15. Ekstrand, Michael D., John T. Riedl, and Joseph A. Konstan. "Collaborative filtering recommender systems." Foundations and Trends in Human-Computer Interaction 4.2,2011,pp. 81-173.

16. 張孫浩,網路評比資料之統計分析,國立政治大學統計學系碩士論文,2011

17. Bottou, Léon. "Stochastic gradient descent tricks." Neural Networks: Tricks of the Trade. Springer Berlin Heidelberg, 2012,pp. 421-436.

18. Zhuang, Yong, et al. "A fast parallel sgd for matrix factorization in shared memory systems." Proceedings of the 7th ACM conference on Recommender systems. ACM, 2013.

19. 張良卉,矩陣分解法對網路評比資料分析之探討,國立政治大學統計學系碩士論文,2013

20. Fang, Xiaowen. A Study of Recommender Systems with Applications. Diss. UNIVERSITY OF MINNESOTA,2014.

21. GroupLens Research. Retrieved JUN,2015,from http://www.grouplens.org		zh_TW