Adaptive ADMM with Spectral Penalty Parameter Selection
About
The alternating direction method of multipliers (ADMM) is a versatile tool for solving a wide range of constrained optimization problems, with differentiable or non-differentiable objective functions. Unfortunately, its performance is highly sensitive to a penalty parameter, which makes ADMM often unreliable and hard to automate for a non-expert user. We tackle this weakness of ADMM by proposing a method to adaptively tune the penalty parameters to achieve fast convergence. The resulting adaptive ADMM (AADMM) algorithm, inspired by the successful Barzilai-Borwein spectral method for gradient descent, yields fast convergence and relative insensitivity to the initial stepsize and problem scaling.
Zheng Xu, Mario A. T. Figueiredo, Tom Goldstein• 2016
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Elastic net regression | Synthetic Elastic net regression | Runtime (s)0.046 | 69 | |
| Sparse CT Image Reconstruction | l1 fidelity TV reg | Relative Residual0.496 | 14 | |
| Constrained Quadratic Optimization | Scaled Quads m = 2 | Relative Residual0.409 | 14 | |
| Constrained Quadratic Optimization | Scaled Quads m = 0 | Relative Residual1.52e-5 | 14 | |
| Constrained Quadratic Optimization | Scaled Quads m = 1 | Relative Residual0.0074 | 7 | |
| Sum of quadratics with multiple constraints optimization | Scaled Quads m = 1 | Median Relative Residual0.0051 | 7 | |
| Constrained Quadratic Optimization | Complex Quads | Relative residual8.29e-9 | 7 | |
| Support Vector Machine | synthetic 1 | Iterations19 | 5 | |
| Elastic net regression | MNIST | Iterations40 | 5 | |
| Elastic net regression | RCV1 | Iterations20 | 5 |
Showing 10 of 51 rows