Byzantine Resilience With Reputation Scores
Jayanth Reddy Regatti, Hao Chen, Abhishek Gupta · 2022
Training distributed machine learning algorithms is prone to Byzantine attacks where the adversarial workers send corrupted model updates to derail the training. In this paper, we propose a reputation score based gradient aggregation as a possible solution. We introduce two novel stochastic gradient descent algorithms, ByGARS (Byzantine Gradient Aggregation using Reputation Scores) and By$\mathbf{GARS++}$that involve computing reputation scores (of workers) using an auxiliary dataset at the server. These reputation scores are then used for aggregating the gradients (model updates) at the server. Under reasonable assumptions, we show that using these reputation scores is robust to any number of adversaries and prove the convergence of By$\mathbf{GARS++}$for strongly convex objective functions using results from two-timescale stochastic approximation theory. The computational complexity of By$\mathbf{GARS++}$is the same as the usual distributed stochastic gradient descent method with only an additional inner product computation in every iteration. We also demonstrate the effectiveness of the algorithms for non-convex learning problems using MNIST and CIFAR-10 datasets against almost all state-of-the-art Byzantine attacks.