New Techniques for Algorithm Portfolio Design

TL;DR

This paper introduces a novel approach to algorithm portfolio design that combines scheduling and machine learning techniques, providing theoretical guarantees and improving performance across multiple AI problem domains.

Contribution

It presents a new integrated technique addressing both scheduling and machine learning aspects of algorithm portfolios, with theoretical guarantees and practical improvements.

Findings

Improved performance on SAT, integer programming, and planning tasks.

The proposed method offers theoretical guarantees.

Demonstrated effectiveness over existing approaches.

Abstract

We present and evaluate new techniques for designing algorithm portfolios. In our view, the problem has both a scheduling aspect and a machine learning aspect. Prior work has largely addressed one of the two aspects in isolation. Building on recent work on the scheduling aspect of the problem, we present a technique that addresses both aspects simultaneously and has attractive theoretical guarantees. Experimentally, we show that this technique can be used to improve the performance of state-of-the-art algorithms for Boolean satisfiability, zero-one integer programming, and A.I. planning.