FedWS: Uma Nova Abordagem para Aprendizado Federado usando Dados Heterogêneos
Resumo
Drones suportam diversos tipos de aplicações envolvendo Sistemas Inteligentes de Transporte que visam entregar serviços seguros, sustentáveis e autônomos. Para isso, o Aprendizado Federado (Federated Learning - FL) provê aprendizado de máquina distribuído e privacidade para esses serviços. Entretanto, FL pode ser afetado negativamente pela heterogeneidade dos dados dos clientes. Neste artigo, é proposta uma técnica de FL para dados heterogêneos que suaviza localmente os pesos das camadas convolucionais de uma rede neural, com o objetivo de melhorar o resultado das rodadas de aprendizado. Os resultados mostram que a técnica proposta obteve uma acurácia com 3% à 6% superior e uma redução no custo de comunicação de 25% à 50% comparado as outras técnicas em tarefas de classificação de imagens no dataset EuroSAT.
Referências
Asad, M., Moustafa, A., Ito, T., and Aslam, M. (2021). Evaluating the communication efficiency in federated learning algorithms. In 24th IEEE International Conference on Computer Supported Cooperative Work in Design (CSCWD), pages 552–557.
Butler, L., Yigitcanlar, T., and Paz, A. (2020). Smart urban mobility innovations: A comprehensive review and evaluation. IEEE ACCESS, 8:196034–196049.
Causa, F., Franzone, A., and Fasano, G. (2023). Strategic and tactical path planning for urban air mobility: Overview and application to real-world use cases. Drones, 7(1):11.
Cohen, A. P., Shaheen, S. A., and Farrar, E. M. (2021). Urban air mobility: History, ecosystem, market potential, and challenges. IEEE Transactions on Intelligent Transportation Systems, 22(9):6074–6087.
Du, Z., Sun, J., Li, A., Chen, P.-Y., Zhang, J., Li, H. H., and Chen, Y. (2022). Rethinking normalization methods in federated learning. In 3rd International Workshop on Distributed Machine Learning, pages 16–22.
Helber, P., Bischke, B., Dengel, A., and Borth, D. (2019). Eurosat: A novel dataset and deep learning benchmark for land use and land cover classification. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 12(7):2217–2226.
Ioffe, S. and Szegedy, C. (2015). Batch normalization: Accelerating deep network training by reducing internal covariate shift. In International Conference on Machine Learning (ICML), pages 448–456.
LeCun, Y., Bengio, Y., and Hinton, G. (2015). Deep learning. Nature, 521(7553):436–444.
Li, X., Jiang, M., Zhang, X., Kamp, M., and Dou, Q. (2021). Fedbn: Federated learning on non-iid features via local batch normalization. arXiv preprint arXiv:2102.07623.
Liu, S., Yu, J., Deng, X., and Wan, S. (2021). Fedcpf: An efficient-communication federated learning approach for vehicular edge computing in 6g communication networks. IEEE Transactions on Intelligent Transportation Systems, 23(2):1616–1629.
Ma, X., Zhu, J., Lin, Z., Chen, S., and Qin, Y. (2022). A state-of-the-art survey on solving non-iid data in federated learning. Future Generation Computer Systems, 135:244–258.
McMahan, B., Moore, E., Ramage, D., Hampson, S., and y Arcas, B. A. (2017). Communication-efficient learning of deep networks from decentralized data. In Artificial intelligence and statistics, pages 1273–1282.
Qiao, S., Wang, H., Liu, C., Shen, W., and Yuille, A. (2019). Micro-batch training with batch-channel normalization and weight standardization. arXiv preprint arXiv:1903.10520.
Wu, Y. and He, K. (2018). Group normalization. In European Conference on Computer Vision (ECCV), pages 3–19.
Yu, F., Zhang, W., Qin, Z., Xu, Z., Wang, D., Liu, C., Tian, Z., and Chen, X. (2021). Fed2: Feature-aligned federated learning. In 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 2066–2074.
Zhao, Y., Li, M., Lai, L., Suda, N., Civin, D., and Chandra, V. (2018). Federated learning with non-iid data. arXiv preprint arXiv:1806.00582.
Zhu, H., Xu, J., Liu, S., and Jin, Y. (2021). Federated learning on non-iid data: A survey. Neurocomputing, 465:371–390.
Butler, L., Yigitcanlar, T., and Paz, A. (2020). Smart urban mobility innovations: A comprehensive review and evaluation. IEEE ACCESS, 8:196034–196049.
Causa, F., Franzone, A., and Fasano, G. (2023). Strategic and tactical path planning for urban air mobility: Overview and application to real-world use cases. Drones, 7(1):11.
Cohen, A. P., Shaheen, S. A., and Farrar, E. M. (2021). Urban air mobility: History, ecosystem, market potential, and challenges. IEEE Transactions on Intelligent Transportation Systems, 22(9):6074–6087.
Du, Z., Sun, J., Li, A., Chen, P.-Y., Zhang, J., Li, H. H., and Chen, Y. (2022). Rethinking normalization methods in federated learning. In 3rd International Workshop on Distributed Machine Learning, pages 16–22.
Helber, P., Bischke, B., Dengel, A., and Borth, D. (2019). Eurosat: A novel dataset and deep learning benchmark for land use and land cover classification. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 12(7):2217–2226.
Ioffe, S. and Szegedy, C. (2015). Batch normalization: Accelerating deep network training by reducing internal covariate shift. In International Conference on Machine Learning (ICML), pages 448–456.
LeCun, Y., Bengio, Y., and Hinton, G. (2015). Deep learning. Nature, 521(7553):436–444.
Li, X., Jiang, M., Zhang, X., Kamp, M., and Dou, Q. (2021). Fedbn: Federated learning on non-iid features via local batch normalization. arXiv preprint arXiv:2102.07623.
Liu, S., Yu, J., Deng, X., and Wan, S. (2021). Fedcpf: An efficient-communication federated learning approach for vehicular edge computing in 6g communication networks. IEEE Transactions on Intelligent Transportation Systems, 23(2):1616–1629.
Ma, X., Zhu, J., Lin, Z., Chen, S., and Qin, Y. (2022). A state-of-the-art survey on solving non-iid data in federated learning. Future Generation Computer Systems, 135:244–258.
McMahan, B., Moore, E., Ramage, D., Hampson, S., and y Arcas, B. A. (2017). Communication-efficient learning of deep networks from decentralized data. In Artificial intelligence and statistics, pages 1273–1282.
Qiao, S., Wang, H., Liu, C., Shen, W., and Yuille, A. (2019). Micro-batch training with batch-channel normalization and weight standardization. arXiv preprint arXiv:1903.10520.
Wu, Y. and He, K. (2018). Group normalization. In European Conference on Computer Vision (ECCV), pages 3–19.
Yu, F., Zhang, W., Qin, Z., Xu, Z., Wang, D., Liu, C., Tian, Z., and Chen, X. (2021). Fed2: Feature-aligned federated learning. In 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 2066–2074.
Zhao, Y., Li, M., Lai, L., Suda, N., Civin, D., and Chandra, V. (2018). Federated learning with non-iid data. arXiv preprint arXiv:1806.00582.
Zhu, H., Xu, J., Liu, S., and Jin, Y. (2021). Federated learning on non-iid data: A survey. Neurocomputing, 465:371–390.
Publicado
06/08/2023
Como Citar
VIEIRA, Flávio; CAMPOS, Carlos Alberto V..
FedWS: Uma Nova Abordagem para Aprendizado Federado usando Dados Heterogêneos. In: WORKSHOP EM DESEMPENHO DE SISTEMAS COMPUTACIONAIS E DE COMUNICAÇÃO (WPERFORMANCE), 22. , 2023, João Pessoa/PB.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2023
.
p. 1-12.
ISSN 2595-6167.
DOI: https://doi.org/10.5753/wperformance.2023.230814.
