Publications-Theses
Article View/Open
Publication Export
-
題名 以類神經網路解決情境式推薦問題
A Neural Network Approach to the Contextual-Bandit Problem作者 陳高欽
Chen, Kao-Chin貢獻者 林怡伶<br>蕭舜文
Lin, Yi-Ling<br>Hsiao, Shun-Wen
陳高欽
Chen, Kao-Chin關鍵詞 情境式推薦
多選項推薦
神經網路
推薦系統
contextual bandit
multi-armed bandit
neural network
recommendation system日期 2020 上傳時間 2-Sep-2020 11:46:11 (UTC+8) 摘要 為了向用戶提供合適的商品推薦，推薦系統已在市場中廣泛應用。儘管市場上衝刺著各種可以使用的數據分析，但是冷啟動（Cold Start）問題對於新進用戶來說仍然是一個大問題。許多最新的推薦算法在假設用戶和商品保持線性關係的前提下設計了演算法，而實際上大多數情況下兩者間存在非線性關係。這項研究開發了一種使用神經網絡（NN）和情境式是推薦的演算法來處理非線性特徵和探索利用的權衡。推薦系統可以有效地預測新進用戶的喜好，還可以快速探索快速變化的喜好。通過將貝葉斯網絡（Bayesian networks）和自動編碼器（AE）集成到NN中，我們的系統, NN Contextual Bandit（NNCB）可以利用不同程度的探索和開發。因此，我們的系統能快速適應情境的變化。我們採用真實世界中的影片評分數據集來證明所提出系統的有效性，與傳統的情境是推薦演算法相比，該系統大約4％的優於就演算法。
Recommendations have been wildly applied in marketplaces to provide right items to users. While various heterogeneous data available in marketplaces, the cold start problem is still a big issue for newcomers. Many state-of-the-art recommendation algorithms were designed on the assumption that users and items remain a linear relationship, while most cases exist nonlinear relationship in reality. This study develops an algorithm using neural network (NN) and contextual bandit to deal with nonlinear context and explore-exploit tradeoff. The recommendation system could effectively predict newcomers’ preferences and also provide quick exploration for fast- changing preferences. By integrating Bayesian networks and AutoEncoder (AE) in the NN, our system, NN Contextual Bandit (NNCB), could leverage different levels of exploration and exploitation. Thus, the proposed recommendation can quickly adapt to the real-time context. We adopt real-world video rating dataset to demonstrate the effectiveness of the proposed system which improve 4% regret as the conventional bandit algorithms.參考文獻 Akter, S., & Wamba, S. F. (2016). Big data analytics in E-commerce: a systematic review and agenda for future research. Electronic Markets, 26(2), 173–194. https://doi.org/10.1007/s12525-016-0219-0Auer, P., & Ortner, R. (2010). UCB revisited: Improved regret bounds for the stochastic multi-armed bandit problem. Periodica Mathematica Hungarica, 61(1), 55–65. https://doi.org/10.1007/s10998-010-3055-6Blundell, C., Cornebise, J., Kavukcuoglu, K., & Wierstra, D. (2015). Weight Uncertainty in Neural Networks. Proceedings of the 32nd International Conference on Machine Learning, Lille, France, 2015. JMLR: W&CP Volume 37. Copy- Right 2015 by the Author(S)., 37. https://doi.org/10.1002/etc.712Chu, W., Li, L., Reyzin, L., & Schapire, R. E. (2011). Contextual bandits with linear Payoff functions. Journal of Machine Learning Research, 15, 208–214.Kingma, D. P., & Welling, M. (2013). Auto-encoding variational bayes. ArXiv Preprint ArXiv:1312.6114.Kiron, D., Prentice, P. K., & Ferguson, R. B. (2014). The Analytics Mandate. MIT Sloan Management Review, 55(4), 1.Li, L., Chu, W., Langford, J., & Schapire, R. E. (2010). A contextual-bandit approach to personalized news article recommendation. Proceedings of the 19th International Conference on World Wide Web, WWW ’10, 661–670. https://doi.org/10.1145/1772690.1772758Liu, B., Wei, Y., Zhang, Y., Yan, Z., & Yang, Q. (2018). Transferable Contextual Bandit for Cross-Domain Recommendation. Aaai, 3619–3626.Schafer, J. Ben, Konstan, J. A., & Riedl, J. (2001). E-commerce recommendation applications. Data Mining and Knowledge Discovery, 5(1–2), 115–153.Sivapalan, S., Sadeghian, A., Rahnama, H., & Madni, A. M. (2014). Recommender systems in e-commerce. World Automation Congress Proceedings, 179–184. https://doi.org/10.1109/WAC.2014.6935763Slivkins, A. (2019). Introduction to Multi-Armed Bandits. (January 2017). Retrieved from http://arxiv.org/abs/1904.07272Zhou, L., & Brunskill, E. (2016). Latent contextual bandits and their application to personalized recommendations for new users. IJCAI International Joint Conference on Artificial Intelligence, 2016-Janua, 3646–3653. 描述 碩士
國立政治大學
資訊管理學系
107356016資料來源 http://thesis.lib.nccu.edu.tw/record/#G0107356016 資料類型 thesis dc.contributor.advisor 林怡伶<br>蕭舜文 zh_TW dc.contributor.advisor Lin, Yi-Ling<br>Hsiao, Shun-Wen en_US dc.contributor.author (Authors) 陳高欽 zh_TW dc.contributor.author (Authors) Chen, Kao-Chin en_US dc.creator (作者) 陳高欽 zh_TW dc.creator (作者) Chen, Kao-Chin en_US dc.date (日期) 2020 en_US dc.date.accessioned 2-Sep-2020 11:46:11 (UTC+8) - dc.date.available 2-Sep-2020 11:46:11 (UTC+8) - dc.date.issued (上傳時間) 2-Sep-2020 11:46:11 (UTC+8) - dc.identifier (Other Identifiers) G0107356016 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/131493 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 資訊管理學系 zh_TW dc.description (描述) 107356016 zh_TW dc.description.abstract (摘要) 為了向用戶提供合適的商品推薦，推薦系統已在市場中廣泛應用。儘管市場上衝刺著各種可以使用的數據分析，但是冷啟動（Cold Start）問題對於新進用戶來說仍然是一個大問題。許多最新的推薦算法在假設用戶和商品保持線性關係的前提下設計了演算法，而實際上大多數情況下兩者間存在非線性關係。這項研究開發了一種使用神經網絡（NN）和情境式是推薦的演算法來處理非線性特徵和探索利用的權衡。推薦系統可以有效地預測新進用戶的喜好，還可以快速探索快速變化的喜好。通過將貝葉斯網絡（Bayesian networks）和自動編碼器（AE）集成到NN中，我們的系統, NN Contextual Bandit（NNCB）可以利用不同程度的探索和開發。因此，我們的系統能快速適應情境的變化。我們採用真實世界中的影片評分數據集來證明所提出系統的有效性，與傳統的情境是推薦演算法相比，該系統大約4％的優於就演算法。 zh_TW dc.description.abstract (摘要) Recommendations have been wildly applied in marketplaces to provide right items to users. While various heterogeneous data available in marketplaces, the cold start problem is still a big issue for newcomers. Many state-of-the-art recommendation algorithms were designed on the assumption that users and items remain a linear relationship, while most cases exist nonlinear relationship in reality. This study develops an algorithm using neural network (NN) and contextual bandit to deal with nonlinear context and explore-exploit tradeoff. The recommendation system could effectively predict newcomers’ preferences and also provide quick exploration for fast- changing preferences. By integrating Bayesian networks and AutoEncoder (AE) in the NN, our system, NN Contextual Bandit (NNCB), could leverage different levels of exploration and exploitation. Thus, the proposed recommendation can quickly adapt to the real-time context. We adopt real-world video rating dataset to demonstrate the effectiveness of the proposed system which improve 4% regret as the conventional bandit algorithms. en_US dc.description.tableofcontents CHAPTER 1 INTRODUCTION 1CHAPTER 2 LITERATURE REVIEW 32-1 Multi-Armed Bandits (MAB) 32-2 Contextual Bandits 3CHAPTER 3 The Proposed Framework 53-1 Design 53-2 NNCB using Bayes by Backprop (NNCB-BNN) 63-3 NNCB using VAE (NNCB-VAE) 93-4 Implementation Environment 11CHAPTER 4 Experimental Results 124-1 Dataset 124-2 Comparison between Different Algorithms 134-3 Dealing with Sorted Data 144-4 Dealing with Balanced Data 17CHAPTER 5 Conclusion and Future Work 19REFERENCE 21 zh_TW dc.format.extent 1100997 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0107356016 en_US dc.subject (關鍵詞) 情境式推薦 zh_TW dc.subject (關鍵詞) 多選項推薦 zh_TW dc.subject (關鍵詞) 神經網路 zh_TW dc.subject (關鍵詞) 推薦系統 zh_TW dc.subject (關鍵詞) contextual bandit en_US dc.subject (關鍵詞) multi-armed bandit en_US dc.subject (關鍵詞) neural network en_US dc.subject (關鍵詞) recommendation system en_US dc.title (題名) 以類神經網路解決情境式推薦問題 zh_TW dc.title (題名) A Neural Network Approach to the Contextual-Bandit Problem en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) Akter, S., & Wamba, S. F. (2016). Big data analytics in E-commerce: a systematic review and agenda for future research. Electronic Markets, 26(2), 173–194. https://doi.org/10.1007/s12525-016-0219-0Auer, P., & Ortner, R. (2010). UCB revisited: Improved regret bounds for the stochastic multi-armed bandit problem. Periodica Mathematica Hungarica, 61(1), 55–65. https://doi.org/10.1007/s10998-010-3055-6Blundell, C., Cornebise, J., Kavukcuoglu, K., & Wierstra, D. (2015). Weight Uncertainty in Neural Networks. Proceedings of the 32nd International Conference on Machine Learning, Lille, France, 2015. JMLR: W&CP Volume 37. Copy- Right 2015 by the Author(S)., 37. https://doi.org/10.1002/etc.712Chu, W., Li, L., Reyzin, L., & Schapire, R. E. (2011). Contextual bandits with linear Payoff functions. Journal of Machine Learning Research, 15, 208–214.Kingma, D. P., & Welling, M. (2013). Auto-encoding variational bayes. ArXiv Preprint ArXiv:1312.6114.Kiron, D., Prentice, P. K., & Ferguson, R. B. (2014). The Analytics Mandate. MIT Sloan Management Review, 55(4), 1.Li, L., Chu, W., Langford, J., & Schapire, R. E. (2010). A contextual-bandit approach to personalized news article recommendation. Proceedings of the 19th International Conference on World Wide Web, WWW ’10, 661–670. https://doi.org/10.1145/1772690.1772758Liu, B., Wei, Y., Zhang, Y., Yan, Z., & Yang, Q. (2018). Transferable Contextual Bandit for Cross-Domain Recommendation. Aaai, 3619–3626.Schafer, J. Ben, Konstan, J. A., & Riedl, J. (2001). E-commerce recommendation applications. Data Mining and Knowledge Discovery, 5(1–2), 115–153.Sivapalan, S., Sadeghian, A., Rahnama, H., & Madni, A. M. (2014). Recommender systems in e-commerce. World Automation Congress Proceedings, 179–184. https://doi.org/10.1109/WAC.2014.6935763Slivkins, A. (2019). Introduction to Multi-Armed Bandits. (January 2017). Retrieved from http://arxiv.org/abs/1904.07272Zhou, L., & Brunskill, E. (2016). Latent contextual bandits and their application to personalized recommendations for new users. IJCAI International Joint Conference on Artificial Intelligence, 2016-Janua, 3646–3653. zh_TW dc.identifier.doi (DOI) 10.6814/NCCU202001541 en_US
