Skip to content
Book Open access

Interpretable Algorithm Selection via Multi-Objective Evolutionary Decision Trees

Jul 2026 · Annual Conference on Genetic and Evolutionary Computation · 0 citations · 22 references
Computer Science

TL;DR

MODT-ASP, a multi-objective evolutionary framework designed to construct interpretable decision trees for ASP overcomes the limitations of existing methods, by employing a customized encoding scheme and specialized genetic operators to explore trade-offs between predictive accuracy and model complexity via Pareto optimization.

Abstract

The Algorithm Selection Problem (ASP) aims to identify the most suitable algorithm for a given problem instance. This paper introduces MODT-ASP, a multi-objective evolutionary framework designed to construct interpretable decision trees for ASP. The proposed approach overcomes the limitations of existing methods, such as scalability constraints and the challenge of balancing competing objectives, by employing a customized encoding scheme and specialized genetic operators. These components effectively explore trade-offs between predictive accuracy and model complexity via Pareto optimization. An enhanced version, EMODT-ASP, further integrates refined control mechanisms to improve generalization. Comprehensive experimental evaluations demonstrate the robustness and effectiveness of the proposed framework. In a large-scale linear programming benchmark comprising 1,004 problems and 532 algorithms, MODT-ASP consistently produces high-quality Pareto-optimal solutions. Furthermore, in the Open Algorithm Selection Challenge (OASC), evaluated across 8 heterogeneous scenarios, the proposed approach achieves third place overall, outperforming 6 of 8 OASC competitors and all 3 IP+VND variants reported in the recent literature. These results confirm its strong cross-domain applicability.

Read PDF

Similar papers

Book Open access Jul 2026

Hybridizing Genetic Algorithms with Pareto Local Search for Multi-Objective Rule Learning

This work investigates the hybridization of a Genetic Algorithm with Pareto Local Search to improve the exploration of non-dominated rulesets and provides a documented baseline for hybrid metaheuristics in rule mining.

Evgueni Blanquart, L. Jourdan, Nadarajen Veerapen · 0 citations
Book Open access Jul 2026

QDEvo: A Multi-Objective Quality-Diversity Framework for Automated Heuristic Design

Large Language Models (LLMs) are opening new directions for automated heuristic design (AHD), allowing evolutionary methods to create and enhance heuristics for constrained optimization problems (COPs). However, most existing approaches face the challenge of the exploration-exploitation balance, where the evolution needs to escape convergence to homogeneous populations and discover as large a heuristic landscape as possible. To address this challenge, we introduce Quality-Diversity Evolution (QDEvo), a multi-objective framework that integrates Quality-Diversity optimization with LLM-based AHD. At its core, QDEvo employs a semantic survival selection mechanism that clusters algorithms by functional similarity, then applies local Pareto competition. Evaluation on well-known COPs benchmarks and real-world problems shows that our method consistently outperforms the state-of-the-art baseline in both Hypervolume and Inverted Generational Distance metrics. These results facilitate further exploration of the algorithmic design space, while ensuring competitive solution quality and efficiency.

Nam Do Khanh, Nhat Nguyen Tran Minh, Dat Pham Vu Tuan et al. · 1 citation
Review Open access Jul 2026

From Pareto to Neural: A Mathematical Survey of Multi-Objective Optimization Algorithms—With Applications to Software Testing

A systematic mapping study of multi-objective optimization algorithms, tracing their evolution from classical Pareto-based methods toward AI-driven and hybrid approaches, with software testing as the primary application domain, and outlining a research roadmap for the next generation of multi-objective optimization systems that combine the complementary mathematical strengths of neural function approximation and evolutionary diversity preservation.

Xufan Zheng, Waqas Rasheed · 0 citations
Book Open access Jul 2026

Study of Decomposition and Ants Allocation in Multi-Objective Partial Classification Rule Induction

This work investigates how balancing the number of decomposed subproblems and ants impacts performance under a fixed computational budget and introduces MOEA/D-AM, a hybrid algorithm that adapts this decomposition strategy for partial classification rule induction.

Adel Abdelkader Mokadem, Marie-Éléonore Kessaci, Laetitia Jourdan et al. · 0 citations
Preprint Jul 2026

A new dual-population constrained multi-objective evolutionary optimization algorithm with repair constraint handling for structural optimization

This study introduces a novel constrained multi-objective evolutionary algorithm, termed DPCME, which employs two interacting populations that exchange information, enabling effective global exploration and reducing the risk of convergence to local optima.

Fardad Homafar, Jasmin Jelovica · 0 citations
Book Open access Jul 2026

Benchmarking Metaheuristics for Mixed-Integer Optimization

This paper compares six different strategies identified in a recent survey, categorizing them as either global strategies (GS), which treat the decision vector as a unified type, or decomposition-based strategies (DS), which split the problem into continuous and discrete subproblems.

Thomas Lang, Denis Pallez · 0 citations