DOI QR코드

DOI QR Code

An Implementation of Federated Learning based on Blockchain

블록체인 기반의 연합학습 구현

  • 박준범 (한국항공대학교 컴퓨터 공학과) ;
  • 박종서 (한국항공대학교 컴퓨터 공학과)
  • Received : 2020.08.03
  • Accepted : 2020.08.25
  • Published : 2020.08.30

Abstract

Deep learning using an artificial neural network has been recently researched and developed in various fields such as image recognition, big data and data analysis. However, federated learning has emerged to solve issues of data privacy invasion and problems that increase the cost and time required to learn. Federated learning presented learning techniques that would bring the benefits of distributed processing system while solving the problems of existing deep learning, but there were still problems with server-client system and motivations for providing learning data. So, we replaced the role of the server with a blockchain system in federated learning, and conducted research to solve the privacy and security problems that are associated with federated learning. In addition, we have implemented a blockchain-based system that motivates users by paying compensation for data provided by users, and requires less maintenance costs while maintaining the same accuracy as existing learning. In this paper, we present the experimental results to show the validity of the blockchain-based system, and compare the results of the existing federated learning with the blockchain-based federated learning. In addition, as a future study, we ended the thesis by presenting solutions to security problems and applicable business fields.

인공신경망(artficial neural networks)를 활용한 딥러닝은 최근 이미지인식, 빅데이터 및 데이터분석 등 다양한 분야에서 연구되고 개발이 진행되고 있다. 하지만 데이터 프라이버시 침해 이슈와 학습을 많이 할수록 소모 비용과 시간이 증가하는 문제점이 있어서 이를 해결하기 위해 연합학습(Federated Learning)이 연구되었다. 연합학습에서는 프라이버시 문제를 완화하면서, 분산 처리 시스템의 이점을 가져오는 학습기법을 제시하였다. 하지만 여전히 연합학습에서도 프라이버시 및 보안 문제가 존재한다. 그래서 우리는 연합학습의 서버에 해당하는 부분을 블록체인으로 대체하여 연합학습의 문제점인 프라이버시 문제와 보안 문제를 해결하였다. 또한 사용자가 제출하는 데이터에 대한 보상을 지급하여서 동기를 부여하고, 기존 성능은 유지하면서도 더 적은 비용의 유지비를 필요로 하는 시스템을 연구하였다. 본 논문에서는 우리가 개발한 시스템의의 타당성을 보이기 위해 실험결과를 제시하면서 기존 연합학습과 연구한 블록체인 기반의 연합학습 결과를 비교한다. 또한 향후 연구로 보안문제에 대한 해법과 와 적용 가능한 비즈니스 분야를 제시를 보여주면서 논문을 마무리 하였다.

Keywords

References

  1. Yann LeCun, Yoshua Bengio & Geoffrey Hinton, "Deep Learning", nature, 27 May 2015.
  2. Jakub Konecny, H. Brendan McMahan, Felix X. Yu, Peter Richtarik, Ananda Theertha Suresh, Dave Bacon, "Federated Learning: Strategies for Improving Communication Efficiency", arXiv:161 0.05492v2 [cs.LG] 30 Oct 2017.
  3. Satoshi Nakamoto, "Bitcoin: A Peer-to-Peer Electronic Cash System", Oct. 2008.
  4. Vitalik Buterin, "A next-generation smart contract and decentralized application platform", cryptorating.eu,2014.
  5. W.Gavin, "Ethereum: A secure decentralised generalised transaction ledger," Ethereum Project Yellow Paper, vol. 151, 2014.
  6. D Hankerson, A Menezes, Elliptic curve cryptography, Boston, Springer, 2011.
  7. H.Bredan McMahan, "Communication-Efficient Learning of Deep Networks from Decentralized Data", AISTATS 2017.
  8. H.Bredan McMahan, "Communication-Efficient Learning of Deep Networks from Decentralized Data", AISTATS 2017.
  9. Lightweight Java library for integration with Ethereum clients Available : https://web3j.io/
  10. E. Gawehn, J. A. Hiss, and G. Schneider, "Deep learning in drug discovery," Molecular informatics, vol. 35, no. 1, pp. 3-14, 2016. https://doi.org/10.1002/minf.201501008
  11. A. B. Kurtulmus and K. Daniel, "Trustless machine learning contracts; evaluating and exchanging machine learning models on the ethereum blockchain," arXiv preprint arXiv:1802.10185, 2018.
  12. J.-S. Weng, J. Weng, M. Li, Y. Zhang, and W. Luo, "Deepchain: Auditable and privacy-preserving deep learning with blockchainbased incentive," Cryptology ePrint Archive, Report 2018/679, 2018, https://eprint.iacr.org/2018/679.
  13. A. Bansal, T. Chen, and S. Zhong, "Privacy preserving backpropagation neural network learning over arbitrarily partitioned data," Neural Computing Applications, vol. 20, no. 1, pp. 143-150, 2011. https://doi.org/10.1007/s00521-010-0346-z
  14. A. Kosba, A. Miller, E. Shi, Z. Wen, and C. Papamanthou, "Hawk: The blockchain model of cryptography and privacy-preserving smart contracts," in Security and Privacy (SP), 2016 IEEE Symposium on. IEEE, 2016, pp. 839-858.
  15. Hyesung Kim, Jihong Park, Mehdi Bennis "On-Device Federated Learning via Blockchain and its Latency Analysis", IEE Communications letters 2019.