Byzantine Resilient Non-Convex SVRG with Distributed Batch Gradient Computations

In this work, we consider the distributed stochastic optimization problem of minimizing a non-convex function $f(x) = \mathbb{E}_{\xi \sim \mathcal{D}} f(x; \xi)$ in an adversarial setting, where the individual functions $f(x; \xi)$ can also be potentially non-convex. We assume that at most $\alpha$-fraction of a total of $K$ nodes can be Byzantines. We propose a robust stochastic variance-reduced gradient (SVRG) like algorithm for the problem, where the batch gradients are computed at the worker nodes (WNs) and the stochastic gradients are computed at the server node (SN). For the non-convex optimization problem, we show that we need $\tilde{O}\left( \frac{1}{\epsilon^{5/3} K^{2/3}} + \frac{\alpha^{4/3}}{\epsilon^{5/3}} \right)$ gradient computations on average at each node (SN and WNs) to reach an $\epsilon$-stationary point. The proposed algorithm guarantees convergence via the design of a novel Byzantine filtering rule which is independent of the problem dimension. Importantly, we capture the effect of the fraction of Byzantine nodes $\alpha$ present in the network on the convergence performance of the algorithm.

[1]  Ohad Shamir,et al.  Optimal Distributed Online Prediction Using Mini-Batches , 2010, J. Mach. Learn. Res..

[2]  Lili Su,et al.  Securing Distributed Machine Learning in High Dimensions , 2018, ArXiv.

[3]  Kannan Ramchandran,et al.  Byzantine-Robust Distributed Learning: Towards Optimal Statistical Rates , 2018, ICML.

[4]  Indranil Gupta,et al.  Zeno: Distributed Stochastic Gradient Descent with Suspicion-based Fault-tolerance , 2018, ICML.

[5]  Alexander J. Smola,et al.  Parallelized Stochastic Gradient Descent , 2010, NIPS.

[6]  Michael I. Jordan,et al.  Non-convex Finite-Sum Optimization Via SCSG Methods , 2017, NIPS.

[7]  Kannan Ramchandran,et al.  Defending Against Saddle Point Attack in Byzantine-Robust Distributed Learning , 2018, ICML.

[8]  Rachid Guerraoui,et al.  Machine Learning with Adversaries: Byzantine Tolerant Gradient Descent , 2017, NIPS.

[9]  Michael I. Jordan,et al.  Less than a Single Pass: Stochastically Controlled Stochastic Gradient , 2016, AISTATS.

[10]  Lili Su,et al.  Securing Distributed Gradient Descent in High Dimensional Statistical Learning , 2018, Proc. ACM Meas. Anal. Comput. Syst..

[11]  Stephen J. Wright,et al.  Hogwild: A Lock-Free Approach to Parallelizing Stochastic Gradient Descent , 2011, NIPS.

[12]  Peng Jiang,et al.  A Linear Speedup Analysis of Distributed Deep Learning with Sparse and Quantized Communication , 2018, NeurIPS.

[13]  Rong Jin,et al.  On the Linear Speedup Analysis of Communication Efficient Momentum SGD for Distributed Non-Convex Optimization , 2019, ICML.

[14]  Qing Ling,et al.  RSA: Byzantine-Robust Stochastic Aggregation Methods for Distributed Learning from Heterogeneous Datasets , 2018, AAAI.

[15]  Alexander J. Smola,et al.  Stochastic Variance Reduction for Nonconvex Optimization , 2016, ICML.

[16]  Dan Alistarh,et al.  Byzantine Stochastic Gradient Descent , 2018, NeurIPS.

[17]  Indranil Gupta,et al.  Phocas: dimensional Byzantine-resilient stochastic gradient descent , 2018, ArXiv.

[18]  Seunghak Lee,et al.  More Effective Distributed ML via a Stale Synchronous Parallel Parameter Server , 2013, NIPS.