Jul 2026· Proceedings of the Genetic and Evolutionary Computation Conference Companion· 0 citations· 25 references
TL;DR
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.
Abstract
Multi-objective evolutionary algorithms based on decomposition (MOEA/D) have established a dominant optimization paradigm successfully extended using Ant Colony Optimization (ACO). However, a rigid architectural assumption within this hybrid framework is the strict assignment of a single ant to evaluate each scalar subproblem, leaving the optimal distribution of computational resources largely unexplored. To address this, we introduce MOEA/D-AM, a hybrid algorithm that adapts this decomposition strategy for partial classification rule induction. By formulating rule induction to optimize multiple objectives simultaneously, we investigate how balancing the number of decomposed subproblems and ants impacts performance under a fixed computational budget. We evaluated five distinct resource configurations across nine benchmark datasets using two global storage strategies, a Pareto archive and a Memory archive. Our analysis reveals significant behavioral differences between these strategies under varying configurations. The findings highlight that the architectural distribution of subproblems and swarm resources fundamentally dictates the model's predictive capabilities. By tuning this trade-off, practitioners can tailor the rule-based classifier to the specific error costs of their domain.
Task scheduling in multi-core systems is a critical NP-hard optimization problem that significantly impacts system performance and resource utilization. This paper proposes a novel hybrid approach combining Bee Colony Optimization (BCO) and Genetic Algorithm (GA) for efficient task scheduling in multi-core processor systems. The hybrid BCO-GA algorithm leverages the global exploration capabilities of BCO and the exploitation strengths of GA to achieve optimal task-to-core assignments while minimizing makespan and balancing system load. The proposed approach incorporates adaptive parameter tuning, elite preservation strategies, and dynamic population management to enhance convergence speed and solution quality. Experimental evaluation using standard benchmark task graphs demonstrates that the hybrid BCO-GA algorithm achieves an average makespan reduction of 18.7% compared to standalone BCO, 15.3% compared to pure GA, and 23.4% compared to the Heterogeneous Earliest Finish Time (HEFT) algorithm. The results also show improved load balancing with a 21.5% reduction in load imbalance factor and 16.8% enhancement in processor utilization. The proposed hybrid approach demonstrates superior performance in handling various task graph characteristics including different Communication-to-Computation Ratios (CCR), task counts, and dependency structures, making it a robust solution for multi-core task scheduling problems.
Igiri C. G, Victor Peters, Igu Ajumoke Elizabeth· Journal of Artificial Intell...· 0 citations
The MOEA/D is the most popular decomposition-based evolutionary algorithm to solve multi-objective optimization problems. However, among the two common decomposition approaches, weighted-sum and Tchebycheff, the existing theoretical research almost exclusively focuses on the latter one. In this first complete mathematical runtime analysis for the MOEA/D using the original weighted-sum decomposition, we show that this variant of the algorithm solves the classic OneMinMax benchmark considerably faster than both the MOEA/D with Tchebycheff decomposition and many other classic algorithms such as the NSGA-II, NSGA-III, SMS-EMOA, and SPEA2. More precisely, we show that already a logarithmic number of subproblems suffices for the algorithm to be efficient, and then typically O(n log2 n) function evaluations suffice to compute the full Pareto front. This beats the other algorithms by a factor of Θ(n/log n). For a second benchmark, the OneJumpZeroJump problem, we show a speed-up by a factor of Θ(n). Overall, this work shows that a further development of the weighted-sum approach might be fruitful. This paper summarizes the work Danyang Zhang, Zerong Zhong, Weijie Zheng, and Benjamin Doerr. 2026. Superior Runtime Guarantees for the MOEA/D Multi-Objective Optimizer via Weighted-Sum Decomposition. In AAAI Conference on Artificial Intelligence, AAAI 2026. AAAI Press, 37187–37194 [14].
Danyang Zhang, Zerong Zhong, Weijie Zheng et al.· Proceedings of the Genetic a...· 0 citations
Multi-objective optimization provides the mathematical foundation for reasoning about trade-offs in complex decision problems, from engineering design to resource allocation. Software testing exemplifies such problems: practitioners must simultaneously optimize for fault detection capability, code coverage, execution cost, and test suite diversity—objectives that are fundamentally incommensurable. Since the early 2000s, multi-objective evolutionary algorithms (MOEAs) such as NSGA-II, MOEA/D, and their many-objective extensions (MOSA; DynaMOSA) have served as the dominant mathematical framework for navigating these trade-offs through Pareto-front approximation with hand-crafted fitness functions. However, the recent emergence of reinforcement learning (RL) and large language models (LLMs) is shifting the optimization paradigm from numerical Pareto-front approximation toward neural, semantically aware decision making over learned representations. This paper presents 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. We survey 120+ papers published from 2000 to 2025 and propose a novel five-level taxonomy (L1–L5) that classifies optimization approaches along the intelligence spectrum: classical MOEAs, ML-guided MOEAs, RL-driven optimization, LLM-driven optimization, and hybrid neuro-evolutionary systems. For each level, we analyze the mathematical problem formulations (Pareto optimality conditions, Markov decision processes, and neural loss landscapes), objective function design, algorithmic convergence properties, and computational complexity. We further conduct a cross-cutting mathematical analysis comparing these paradigms along dimensions of convergence, diversity, scalability, and interpretability. Our survey identifies critical open mathematical challenges: the lack of formal convergence guarantees for LLM-driven optimization, the under-exploration of many-objective (m≥4) formulations in AI-driven testing, the sample complexity of reinforcement learning for combinatorial test optimization, and the absence of standardized benchmarks with known Pareto-optimal frontiers. We conclude by 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.
In evolutionary algorithms (EAs), static operator configurations often depend on human expertise and may lead to search stagnation. Existing adaptive methods either overlook real-time optimization states or require substantial computational overhead. Recent LLM-based methods support adaptive decision-making; however, applying an LLM to a single population requires the model to resolve conflicting exploration and exploitation objectives, which may reduce the reliability of operator selection. To address this limitation, this paper proposes DEAL, a Dual-Pool Evolutionary Algorithm with LLM-Driven Adaptive Operator Selection. DEAL separates the population into an exploitation pool and an exploration pool, enabling the LLM to pursue different search objectives: refining promising solutions in the exploitation pool and exploring alternative candidate structures in the exploration pool. DEAL further incorporates an offline experience collection module and a structured prompt template that jointly provide operator information, population status, and relevant historical evidence. Benchmark experiments on EUC-2D TSP instances show that DEAL reduces the average optimality gap on most tested categories compared with conventional baselines, suggesting the effectiveness of dual-pool LLM-driven operator scheduling.
Teming Qi, Yishun Liu, Chunhua Yang et al.· 2026 IEEE 27th China Confere...· 0 citations
Facility location decisions directly affect operational costs, service quality, and customer allocation. However, minimising total cost may result in an imbalanced distribution of customers among open facilities, requiring both objectives to be considered simultaneously. This study proposes a novel non-dominated sorting adaptive binary artificial bee colony algorithm with adaptive operator selection, called NSABC, for the bi-objective uncapacitated facility location problem. The first objective minimises facility opening and customer assignment costs, while the second minimises customer allocation imbalance among open facilities. NSABC integrates Pareto-based archiving, smart initialisation, adaptive operator selection, and diversity-preservation mechanisms to generate high-quality and diverse trade-off solutions. Computational experiments on 15 OR-Library CAP benchmark instances evaluate the algorithms using Hypervolume and Inverted Generational Distance as complementary Pareto-front performance indicators, together with paired two-sided Wilcoxon signed-rank tests and Holm correction. NSABC achieves higher mean Hypervolume values on most instances and lower mean IGD values on 14 of the 15 instances. The statistical analysis significantly favours NSABC on 11 instances according to Hypervolume and on 9 instances according to IGD, whereas NSGA-III is significantly favoured on only one instance according to IGD. The performance advantages of NSABC were observed across benchmark instances of different sizes and scales, indicating its effectiveness under varying problem structures. These findings indicate that NSABC is a competitive and statistically supported alternative to NSGA-III for bi-objective facility location problems involving both economic efficiency and balanced customer distribution.
Muhammed Resul Aydın, Mehmet Emin Aydin· Mathematics· 0 citations
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· Proceedings of the Genetic a...· 0 citations