Federated Learning: Strategies for Improving Communication Efficiency

作者: Jakub Konečný , Ananda Theertha Suresh , Dave Bacon , Felix X. Yu , Peter Richtarik

DOI:

关键词: Telecommunications linkFederated learningTraining setComputer scienceDistributed computingFull modelQuantization (signal processing)

摘要: … Federated Learning. For simplicity, we consider synchronized algorithms for Federated Learning … We conducted experiments using Federated Learning to train deep neural networks for …

参考文章(15)
Jakub Konečný, Zheng Qu, Peter Richtárik, Semi-Stochastic Coordinate Descent arXiv: Numerical Analysis. ,(2014)
Xiao Lin, Yuchen Zhang, DiSCO: Distributed Optimization for Self-Concordant Empirical Loss international conference on machine learning. pp. 362- 370 ,(2015)
Jakub Konečný, Daniel Ramage, H. Brendan McMahan, Federated Optimization: Distributed Optimization Beyond the Datacenter arXiv: Learning. ,(2015)
M.G. Rabbat, R.D. Nowak, Quantized incremental algorithms for distributed optimization IEEE Journal on Selected Areas in Communications. ,vol. 23, pp. 798- 808 ,(2005) , 10.1109/JSAC.2005.843546
Jeffrey Dean, Greg Corrado, Rajat Monga, Kai Chen, Matthieu Devin, Mark Mao, Marc'aurelio Ranzato, Andrew Senior, Paul Tucker, Ke Yang, Quoc Le, Andrew Ng, None, Large Scale Distributed Deep Networks neural information processing systems. ,vol. 25, pp. 1223- 1231 ,(2012)
Rami Al-Rfou, Marc Pickett, Javier Snaider, Yun-hsuan Sung, Brian Strope, Ray Kurzweil, Conversational Contextual Cues: The Case of Personalization and History for Response Ranking arXiv: Computation and Language. ,(2016)
Barnabás Póczos, Alexander J. Smola, Jakub Konecný, Sashank J. Reddi, Peter Richtárik, AIDE: Fast and Communication Efficient Distributed Optimization arXiv: Optimization and Control. ,(2016)
Mostafa El Gamal, Lifeng Lai, On Randomized Distributed Coordinate Descent with Quantized Updates arXiv: Machine Learning. ,(2016)
Jakub Konečný, Daniel Ramage, Peter Richtarik, H. Brendan McMahan, Federated Optimization: Distributed Machine Learning for On-Device Intelligence arXiv: Learning. ,(2016)