政大機構典藏-National Chengchi University Institutional Repository(NCCUR):Item 140.119/155451

政大機構典藏-National Chengchi University Institutional Repository(NCCUR):Item 140.119/155451

English | 正體中文 | 简体中文 | Post-Print筆數 : 27 | Items with full text/Total items : 114367/145401 (79%)
Visitors : 53046171 Online Users : 376

RC Version 6.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.

Scope

please add "double quotation mark" for query phrases to get precise results

please goto advance search for comprehansive author search

Adv. Search

Home ‧ Login ‧ Upload ‧ Help ‧ About ‧ Administer

Goto mobile version

政大典藏 > College of Informatics > Department of Computer Science > Theses > Item 140.119/155451

Please use this identifier to cite or link to this item: https://nccur.lib.nccu.edu.tw/handle/140.119/155451

Title:	基於命名式網路架構之個性化聯邦推薦系統 Personalized Federated Recommendation Systems Via Named Data Networking
Authors:	蔡明衡 Tsai, Ming-Heng
Contributors:	蔡子傑 Tsai, Tzu-Chieh 蔡明衡 Tsai, Ming-Heng
Keywords:	推薦系統聯邦學習個性化聯邦學習命名資料網路點對點分散式網路資料隱私 Recommender system Federated Learning Personalized Federated Learning Named Data Networking Peer-to-peer Data Privacy
Date:	2024
Issue Date:	2025-02-04 15:43:41 (UTC+8)
Abstract:	隨著資訊技術和人工智慧的持續進步，個人智慧設備的發展促使資料分析和隱私保護的重要性逐漸增加。聯邦學習作為一種新型的機器學習架構，不僅能夠滿足資料隱私的需求，允許分散的資料保持在原始位置，同時還能進行模型的協同訓練。但是聯邦學習在資料非獨立同分佈（Non-IID）情境下，仍面臨諸多挑戰。而個性化聯邦學習是一種有前景的解決方案。本研究聚焦在個性化學習中的客戶端選擇與多中心聯邦學習上面。本研究以不同網路架構對聯邦學習進行了改良和詳細探討。在 NDN 架構中進行客戶端選擇可以以更少的網路成本達到較好的效能；而在 P2P 多中心聯邦學習中，雖然需要提高網路成本，但能夠獲得更好的結果。將這兩者結合後，可以同時達到更低的網路成本和較佳的推薦效能。此外，本研究還採用了元學習的概念，在面對推薦系統中新加入的用戶時，取得了良好的結果。綜上所述，本研究不僅深入探討了個性化聯邦學習所面臨的各種挑戰，還提出了多種基於網路架構的優化策略，尤其是基於 NDN 網路的客戶端選擇方法，顯著降低聯邦學習的通訊成本並提升推薦效能。這些研究成果對理解和優化聯邦學習具有重要的參考價值，並為未來的研究和實際應用提供了概念性驗證。 With the continuous advancement of information technology and artificial intelligence, the development of personal smart devices has emphasized the growing importance of data analysis and privacy protection. Federated learning, as a novel machine learning framework, not only satisfies data privacy requirements by allowing distributed data to remain at their original locations but also enables collaborative model training. However, federated learning faces numerous challenges, particularly under non-independent and identically distributed (Non-IID) data scenarios. Personalized federated learning emerges as a promising solution to address these challenges. This study focuses on client selection and multi-center federated learning within the scope of personalized learning. This research explores and improves federated learning through various network architectures. Client selection within the NDN architecture achieves better performance with reduced network costs, while P2P multi-center federated learning, though incurring higher network costs, delivers superior results. By combining these two approaches, the system achieves lower network costs and improved recommendation performance. Furthermore, the study adopts the concept of meta-learning, achieving favorable results when addressing newly added users in the recommendation system. In conclusion, this study not only delves into the challenges faced by personalized federated learning but also proposes several optimization strategies based on network architectures, particularly the NDN-based client selection method, to significantly reduce communication costs and enhance recommendation performance. These findings provide valuable insights for understanding and optimizing federated learning and offer conceptual validation for future research and practical applications
Reference:	[1] Konečný, J., McMahan, H. B., Ramage, D., & Richtárik, P. (2016). Federated optimization: Distributed machine learning for on-device intelligence. arXiv preprint arXiv:1610.02527. [2] Yang, Q., Liu, Y., Chen, T., & Tong, Y. (2019). Federated machine learning: Concept and applications. ACM Transactions on Intelligent Systems and Technology (TIST), 10(2), 1-19. [3] McMahan, B., Moore, E., Ramage, D., Hampson, S., & y Arcas, B. A. (2017, April). Communication-efficient learning of deep networks from decentralized data. In Artificial intelligence and statistics (pp. 1273-1282). PMLR. [4] Koren, Y., Bell, R., & Volinsky, C. (2009). Matrix factorization techniques for recommender systems. Computer, 42(8), 30-37. [5] Mohammadi, M. (2020). A new non-negative matrix factorization method to build a recommender system. Journal of Research in Science, Engineering and Technology, 8(2), 6-12. [6] Sun, Z., Xu, Y., Liu, Y., He, W., Kong, L., Wu, F., ... & Cui, L. (2022). A survey on federated recommendation systems. arXiv preprint arXiv:2301.00767. [7] Tan, A. Z., Yu, H., Cui, L., & Yang, Q. (2022). Towards personalized federated learning. IEEE Transactions on Neural Networks and Learning Systems. [8] Zhao, Y., Li, M., Lai, L., Suda, N., Civin, D., & Chandra, V. (2018). Federated learning with non-iid data. arXiv preprint arXiv:1806.00582. [9] Jiang, Y., Konečný, J., Rush, K., & Kannan, S. (2019). Improving federated learning personalization via model agnostic meta learning. arXiv preprint arXiv:1909.12488. [10] Zhang, L., Afanasyev, A., Burke, J., Jacobson, V., Claffy, K. C., Crowley, P., ... & Zhang, B. (2014). Named data networking. ACM SIGCOMM Computer Communication Review, 44(3), 66-73. [11] Amadeo, M., Campolo, C., & Molinaro, A. (2016). NDNe: Enhancing named data networking to support cloudification at the edge. IEEE Communications Letters, 20(11), 2264-2267. [12] Amadeo, M., Campolo, C., Iera, A., Molinaro, A., & Ruggeri, G. (2022, May). Client Discovery and Data Exchange in Edge-based Federated Learning via Named Data Networking. In ICC 2022-IEEE International Conference on Communications (pp. 2990-2995). IEEE. [13] Roy, A. G., Siddiqui, S., Pölsterl, S., Navab, N., & Wachinger, C. (2019). Braintorrent: A peer-to-peer environment for decentralized federated learning. arXiv preprint arXiv:1905.06731. [14] Chai, D., Wang, L., Chen, K., & Yang, Q. (2020). Secure federated matrix factorization. IEEE Intelligent Systems, 36(5), 11-20. [15] Mastorakis, S., Afanasyev, A., & Zhang, L. (2017). On the evolution of ndnSIM: An open-source simulator for NDN experimentation. ACM SIGCOMM Computer Communication Review, 47(3), 19-33. [16] Ni, J., Li, J., & McAuley, J. (2019, November). Justifying recommendations using distantly-labeled reviews and fine-grained aspects. In Proceedings of the 2019 conference on empirical methods in natural language processing and the 9th international joint conference on natural language processing (EMNLP-IJCNLP) (pp. 188-197)
Description:	碩士國立政治大學資訊科學系 110753133
Source URI:	http://thesis.lib.nccu.edu.tw/record/#G0110753133
Data Type:	thesis
Appears in Collections:	[Department of Computer Science ] Theses

Files in This Item:

File	Description	Size	Format
313301.pdf		5841Kb	Adobe PDF	0	View/Open

All items in 政大典藏 are protected by copyright, with all rights reserved.

社群 sharing

著作權政策宣告 Copyright Announcement

1.本網站之數位內容為國立政治大學所收錄之機構典藏，無償提供學術研究與公眾教育等公益性使用，惟仍請適度，合理使用本網站之內容，以尊重著作權人之權益。商業上之利用，則請先取得著作權人之授權。
The digital content of this website is part of National Chengchi University Institutional Repository. It provides free access to academic research and public education for non-commercial use. Please utilize it in a proper and reasonable manner and respect the rights of copyright owners. For commercial use, please obtain authorization from the copyright owner in advance.

2.本網站之製作，已盡力防止侵害著作權人之權益，如仍發現本網站之數位內容有侵害著作權人權益情事者，請權利人通知本網站維護人員(nccur@nccu.edu.tw)，維護人員將立即採取移除該數位著作等補救措施。
NCCU Institutional Repository is made to protect the interests of copyright owners. If you believe that any material on the website infringes copyright, please contact our staff(nccur@nccu.edu.tw). We will remove the work from the repository and investigate your claim.

DSpace Software Copyright © 2002-2004 MIT & Hewlett-Packard / Enhanced by NTU Library IR team Copyright © - Feedback