Robust Accelerated Gradient Methods for Smooth Strongly Convex Functions

We study the trade-offs between convergence rate and robustness to gradient errors in designing a first-order algorithm. We focus on gradient descent and accelerated gradient (AG) methods for minim...