Minimax Excess Risk of First-Order Methods for Statistical Learning with Data-Dependent Oracles
arXiv:2307.04679
Abstract
In this paper, our aim is to analyse the generalization capabilities of first-order methods for statistical learning in multiple, different yet related, scenarios including supervised learning, transfer learning, robust learning and federated learning. To do so, we provide sharp upper and lower bounds for the minimax excess risk of strongly convex and smooth statistical learning when the gradient is accessed through partial observations given by a data-dependent oracle. This novel class of oracles can query the gradient with any given data distribution, and is thus well suited to scenarios in which the training data distribution does not match the target (or test) distribution. In particular, our upper and lower bounds are proportional to the smallest mean square error achievable by gradient estimators, thus allowing us to easily derive multiple sharp bounds in the aforementioned scenarios using the extensive literature on parameter estimation.
arXiv comment22 pages, 0 figures
References in corpus (5)
- Train faster, generalize better: Stability of stochastic gradient descent
- SGD Generalizes Better Than GD (And Regularization Doesn't Help)
- Stability of SGD: Tightness Analysis and Improved Bounds
- Stability vs Implicit Bias of Gradient Methods on Separable Data and Beyond
- Tight Risk Bounds for Gradient Descent on Separable Data