Adaptive Algorithms with Sharp Convergence Rates for Stochastic Hierarchical Optimization
By: Xiaochuan Gong, Jie Hao, Mingrui Liu
Potential Business Impact:
Helps computers solve tough problems without knowing how hard they are.
Hierarchical optimization refers to problems with interdependent decision variables and objectives, such as minimax and bilevel formulations. While various algorithms have been proposed, existing methods and analyses lack adaptivity in stochastic optimization settings: they cannot achieve optimal convergence rates across a wide spectrum of gradient noise levels without prior knowledge of the noise magnitude. In this paper, we propose novel adaptive algorithms for two important classes of stochastic hierarchical optimization problems: nonconvex-strongly-concave minimax optimization and nonconvex-strongly-convex bilevel optimization. Our algorithms achieve sharp convergence rates of $\widetilde{O}(1/\sqrt{T} + \sqrt{\bar{\sigma}}/T^{1/4})$ in $T$ iterations for the gradient norm, where $\bar{\sigma}$ is an upper bound on the stochastic gradient noise. Notably, these rates are obtained without prior knowledge of the noise level, thereby enabling automatic adaptivity in both low and high-noise regimes. To our knowledge, this work provides the first adaptive and sharp convergence guarantees for stochastic hierarchical optimization. Our algorithm design combines the momentum normalization technique with novel adaptive parameter choices. Extensive experiments on synthetic and deep learning tasks demonstrate the effectiveness of our proposed algorithms.
Similar Papers
Nonconvex Decentralized Stochastic Bilevel Optimization under Heavy-Tailed Noises
Machine Learning (CS)
Teaches computers to learn better with messy data.
Stochastic Bilevel Optimization with Heavy-Tailed Noise
Machine Learning (CS)
Teaches computers to learn better from messy data.
On the Convergence of Adam-Type Algorithm for Bilevel Optimization under Unbounded Smoothness
Machine Learning (CS)
Helps AI learn faster on complex tasks.