Timezone: »

Biased Stochastic First-Order Methods for Conditional Stochastic Optimization and Applications in Meta Learning
Yifan Hu · Siqi Zhang · Xin Chen · Niao He

Thu Dec 10 09:00 AM -- 11:00 AM (PST) @ Poster Session 5 #1389

Conditional stochastic optimization covers a variety of applications ranging from invariant learning and causal inference to meta-learning. However, constructing unbiased gradient estimators for such problems is challenging due to the composition structure. As an alternative, we propose a biased stochastic gradient descent (BSGD) algorithm and study the bias-variance tradeoff under different structural assumptions. We establish the sample complexities of BSGD for strongly convex, convex, and weakly convex objectives under smooth and non-smooth conditions. Our lower bound analysis shows that the sample complexities of BSGD cannot be improved for general convex objectives and nonconvex objectives except for smooth nonconvex objectives with Lipschitz continuous gradient estimator. For this special setting, we propose an accelerated algorithm called biased SpiderBoost (BSpiderBoost) that matches the lower bound complexity. We further conduct numerical experiments on invariant logistic regression and model-agnostic meta-learning to illustrate the performance of BSGD and BSpiderBoost.

Author Information

Yifan Hu (University of Illinois at Urbana-Champaign)
Siqi Zhang (University of Illinois at Urbana-Champaign)
Xin Chen (University of Illinois at Urbana-Champaign)
Niao He (ETH Zurich)

More from the Same Authors