Solving Constrained CASH Problems with ADMM
Machine Learning
2020-07-14 v2 Optimization and Control
Machine Learning
Abstract
The CASH problem has been widely studied in the context of automated configurations of machine learning (ML) pipelines and various solvers and toolkits are available. However, CASH solvers do not directly handle black-box constraints such as fairness, robustness or other domain-specific custom constraints. We present our recent approach [Liu, et al., 2020] that leverages the ADMM optimization framework to decompose CASH into multiple small problems and demonstrate how ADMM facilitates incorporation of black-box constraints.
Cite
@article{arxiv.2006.09635,
title = {Solving Constrained CASH Problems with ADMM},
author = {Parikshit Ram and Sijia Liu and Deepak Vijaykeerthi and Dakuo Wang and Djallel Bouneffouf and Greg Bramble and Horst Samulowitz and Alexander G. Gray},
journal= {arXiv preprint arXiv:2006.09635},
year = {2020}
}
Comments
7th ICML Workshop on Automated Machine Learning (2020)