Presentation | 2023-01-19 [Short Paper] A Study of decentralized model training method based on traveling model for P2P Federated Learning Kota Maejima, Takayuki Nishio, Asato Yamazaki, Yuko Hara, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Peer-to-Peer(P2P) Federated Learning is a machine learning method that builds models across clients without sharing training data among clients. When the data distribution is Non-IID(non-Independent and Identically Distributed), the accuracy of the learned model in P2P Federated Learning deteriorates than in centralized machine learning, which gathers datasets in a central server. In this paper, we propose a method to prevent the model accuracy degradation in P2P Federated Learning in the Non-IID setting. In the proposed method, a single model is circulated over the network and trained by each client in turn. By appropriately routing the model based on the label distribution among clients, the model can be well-trained on non-IID data, similarly to when trained on IID data. Our evaluation results show that the proposed method converges faster than baselines, GossipSGD and PDMM-SGD, especially when the data stored by each client is far from the IID data. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Machine Learning / Federated Learning / Non-IID |
Paper # | SeMI2022-75 |
Date of Issue | 2023-01-12 (SeMI) |
Conference Information | |
Committee | SeMI |
---|---|
Conference Date | 2023/1/19(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Naruto grand hotel |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Koji Yamamoto(Kyoto Univ.) |
Vice Chair | Kazuya Monden(Hitachi) / Yasunori Owada(NICT) / Shunsuke Saruwatari(Osaka Univ.) |
Secretary | Kazuya Monden(NTT DOCOMO) / Yasunori Owada(Tokyo Univ. of Agri. and Tech.) / Shunsuke Saruwatari(Osaka Univ.) |
Assistant | Yuki Matsuda(NAIST) / Akihito Taya(Aoyama Gakuin Univ.) / Takeshi Hirai(Osaka Univ.) |
Paper Information | |
Registration To | Technical Committee on Sensor Network and Mobile Intelligence |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | [Short Paper] A Study of decentralized model training method based on traveling model for P2P Federated Learning |
Sub Title (in English) | |
Keyword(1) | Machine Learning |
Keyword(2) | Federated Learning |
Keyword(3) | Non-IID |
1st Author's Name | Kota Maejima |
1st Author's Affiliation | Tokyo Institute of Technology(Tokyo Tech) |
2nd Author's Name | Takayuki Nishio |
2nd Author's Affiliation | Tokyo Institute of Technology(Tokyo Tech) |
3rd Author's Name | Asato Yamazaki |
3rd Author's Affiliation | Tokyo Institute of Technology(Tokyo Tech) |
4th Author's Name | Yuko Hara |
4th Author's Affiliation | Tokyo Institute of Technology(Tokyo Tech) |
Date | 2023-01-19 |
Paper # | SeMI2022-75 |
Volume (vol) | vol.122 |
Number (no) | SeMI-341 |
Page | pp.pp.23-24(SeMI), |
#Pages | 2 |
Date of Issue | 2023-01-12 (SeMI) |