MOSS: Multi-Objective Optimization for Stable Rule Sets

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Brian, Mazumder, Rahul
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916869178392576
author Liu, Brian
Mazumder, Rahul
author_facet Liu, Brian
Mazumder, Rahul
contents We present MOSS, a multi-objective optimization framework for constructing stable sets of decision rules. MOSS incorporates three important criteria for interpretability: sparsity, accuracy, and stability, into a single multi-objective optimization framework. Importantly, MOSS allows a practitioner to rapidly evaluate the trade-off between accuracy and stability in sparse rule sets in order to select an appropriate model. We develop a specialized cutting plane algorithm in our framework to rapidly compute the Pareto frontier between these two objectives, and our algorithm scales to problem instances beyond the capabilities of commercial optimization solvers. Our experiments show that MOSS outperforms state-of-the-art rule ensembles in terms of both predictive performance and stability.
format Preprint
id arxiv_https___arxiv_org_abs_2506_08030
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MOSS: Multi-Objective Optimization for Stable Rule Sets
Liu, Brian
Mazumder, Rahul
Optimization and Control
Machine Learning
We present MOSS, a multi-objective optimization framework for constructing stable sets of decision rules. MOSS incorporates three important criteria for interpretability: sparsity, accuracy, and stability, into a single multi-objective optimization framework. Importantly, MOSS allows a practitioner to rapidly evaluate the trade-off between accuracy and stability in sparse rule sets in order to select an appropriate model. We develop a specialized cutting plane algorithm in our framework to rapidly compute the Pareto frontier between these two objectives, and our algorithm scales to problem instances beyond the capabilities of commercial optimization solvers. Our experiments show that MOSS outperforms state-of-the-art rule ensembles in terms of both predictive performance and stability.
title MOSS: Multi-Objective Optimization for Stable Rule Sets
topic Optimization and Control
Machine Learning
url https://arxiv.org/abs/2506.08030