Learned online policies for the ADMM relaxation parameter improve iteration count and runtime on benchmark quadratic programs while maintaining convergence guarantees for time-varying parameters under mild assumptions.
Learning algorithm hyperparam- eters for fast parametric convex optimization
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
representative citing papers
A GNN is trained to predict adaptive step sizes and weights for distributed ADMM by unrolling a fixed number of iterations and minimizing solution error on a problem class.
citing papers explorer
-
Learning Over-Relaxation Policies for ADMM with Convergence Guarantees
Learned online policies for the ADMM relaxation parameter improve iteration count and runtime on benchmark quadratic programs while maintaining convergence guarantees for time-varying parameters under mild assumptions.
-
Learning to accelerate distributed ADMM using graph neural networks
A GNN is trained to predict adaptive step sizes and weights for distributed ADMM by unrolling a fixed number of iterations and minimizing solution error on a problem class.