Optimizing Simulations with Noise-Tolerant Structured Exploration
ICRA, pp. 2970-2977, 2018.
We propose a simple drop-in noise-tolerant replacement for the standard finite difference procedure used ubiquitously in blackbox optimization. In our approach, parameter perturbation directions are defined by a family of structured orthogonal matrices. We show that at the small cost of computing a Fast Walsh-Hadamard/Fourier Transform (F...More
Full Text (Upload PDF)
PPT (Upload PPT)