Distributed Black-Box Optimization via Error Correcting Codes
This addresses the problem of efficient distributed optimization for machine learning practitioners, though it appears incremental as an extension of evolution strategies.
The paper tackles the problem of straggler-resilient distributed black-box optimization by introducing a framework that uses coded search directions and decoding. The method achieves significant improvements in computation times for black-box adversarial attacks on deep convolutional neural networks.
We introduce a novel distributed derivative-free optimization framework that is resilient to stragglers. The proposed method employs coded search directions at which the objective function is evaluated, and a decoding step to find the next iterate. Our framework can be seen as an extension of evolution strategies and structured exploration methods where structured search directions were utilized. As an application, we consider black-box adversarial attacks on deep convolutional neural networks. Our numerical experiments demonstrate a significant improvement in the computation times.