Alternating Implicit Projected SGD and Its Efficient Variants for Equality-constrained Bilevel Optimization

11/14/2022
by   Quan Xiao, et al.
0

Stochastic bilevel optimization, which captures the inherent nested structure of machine learning problems, is gaining popularity in many recent applications. Existing works on bilevel optimization mostly consider either unconstrained problems or constrained upper-level problems. This paper considers the stochastic bilevel optimization problems with equality constraints both in the upper and lower levels. By leveraging the special structure of the equality constraints problem, the paper first presents an alternating implicit projected SGD approach and establishes the Õ(ϵ^-2) sample complexity that matches the state-of-the-art complexity of ALSET <cit.> for unconstrained bilevel problems. To further save the cost of projection, the paper presents two alternating implicit projection-efficient SGD approaches, where one algorithm enjoys the Õ(ϵ^-2/T) upper-level and O(ϵ^-1.5/T^3/4) lower-level projection complexity with O(T) lower-level batch size, and the other one enjoys Õ(ϵ^-1.5) upper-level and lower-level projection complexity with O(1) batch size. Application to federated bilevel optimization has been presented to showcase the empirical performance of our algorithms. Our results demonstrate that equality-constrained bilevel optimization with strongly-convex lower-level problems can be solved as efficiently as stochastic single-level optimization problems.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
06/25/2021

Tighter Analysis of Alternating Stochastic Gradient Method for Stochastic Nested Problems

Stochastic nested optimization, including stochastic compositional, min-...
research
05/21/2021

Online Statistical Inference for Parameters Estimation with Linear-Equality Constraints

Stochastic gradient descent (SGD) and projected stochastic gradient desc...
research
01/04/2023

First-order penalty methods for bilevel optimization

In this paper we study a class of unconstrained and constrained bilevel ...
research
05/07/2015

Fast Spectral Unmixing based on Dykstra's Alternating Projection

This paper presents a fast spectral unmixing algorithm based on Dykstra'...
research
03/19/2021

Empirical Optimization on Post-Disaster Communication Restoration for Social Equality

Disasters are constant threats to humankind, and beyond losses in lives,...
research
10/05/2020

First-order methods for problems with O(1) functional constraints can have almost the same convergence rate as for unconstrained problems

First-order methods (FOMs) have recently been applied and analyzed for s...
research
01/07/2020

An Efficient Gradient Projection Method for Structural Topology Optimization

This paper presents an efficient gradient projection-based method for st...

Please sign up or login with your details

Forgot password? Click here to reset