Artificial Intelligence, Год журнала: 2024, Номер 336, С. 104207 - 104207
Опубликована: Авг. 24, 2024
Язык: Английский
European Journal of Operational Research, Год журнала: 2024, Номер 321(2), С. 345 - 362
Опубликована: Апрель 9, 2024
In this paper, we review the milestones in development of heuristic methods for optimization over last 50 years. We propose a critical analysis main findings and contributions, mainly from European perspective. Starting with roots area that can be traced back to classical philosophers, follow historical path heuristics metaheuristics field operations research list milestones, up latest proposals hybridize machine learning. pay special attention theories changed our way thinking about problem solving, role played by Journal Operational Research these theories. Our approach emphasizes methodologies their connections related areas, which permits identify potential lines future research.
Язык: Английский
Процитировано
19
European Journal of Operational Research, Год журнала: 2024, Номер 318(2), С. 575 - 591
Опубликована: Июнь 7, 2024
Язык: Английский
Процитировано
7
European Journal of Operational Research, Год журнала: 2024, Номер 319(1), С. 102 - 120
Опубликована: Июль 2, 2024
The Quadratic Knapsack Problem (QKP) is a well-studied combinatorial optimization problem with practical applications in various fields such as finance, logistics, and telecommunications. Despite its longstanding interest, the QKP remains challenging due to strong NP-hardness. Moreover, recent studies have introduced new instances where all existing algorithms failed produce good-quality results. In this paper, we aim address these by proposing novel approach enhance regular value function used dynamic programming (DP) literature. Our proposed method considers contribution of each item not only respect items already selected, but also estimates potential yet be considered. Additionally, introduce propagation technique "remove-and-fill-up" local search procedure further improve solution quality. Through extensive computational experiments, our heuristic algorithm demonstrates superior performance compared heuristics, producing optimal or near-optimal solutions for even most demanding instances. Empirical evidence, supported an automated instance space analysis using unbiased metrics, showcases remarkable improvements achieved, surpassing on average quality up 98%, 77% reduction time.
Язык: Английский
Процитировано
5
ACM Transactions on Evolutionary Learning and Optimization, Год журнала: 2024, Номер 5(1), С. 1 - 19
Опубликована: Июнь 21, 2024
Choosing a set of benchmark problems is often key component any empirical evaluation iterative optimization heuristics. In continuous, single-objective optimization, several sets have become widespread, including the well-established BBOB suite. While this suite designed to enable rigorous benchmarking, it also commonly used for testing methods such as algorithm selection, which was never around. We present MA-BBOB function generator, uses functions in an affine combination. work, we describe full procedure create these combinations and highlight tradeoffs design decisions, specifically choice place optimum uniformly at random domain. then illustrate how generator can be gain more low-level insight into landscapes through use exploratory landscape analysis. Finally, show potential use-case generating wide training data selectors. Using setup, that basic scheme using features predict best does not lead optimal results, selector trained purely on generalizes poorly combinations.
Язык: Английский
Процитировано
4
Expert Systems with Applications, Год журнала: 2024, Номер 255, С. 124484 - 124484
Опубликована: Июнь 18, 2024
We propose a novel Multi-Neighborhood Simulated Annealing approach to address the Capacitated Dispersion Problem. It makes use of three neighborhoods, adapted from similar proposals literature. Our search method, properly engineered and tuned, is able consistently improve state-of-the-art methods on almost all instances public benchmarks. In addition, we highlight limitations current datasets new, more challenging one, obtained by sampling data real maps population density. Finally, two compact mathematical models that obtain good bounds small/medium size as well as, with long runs, large ones.
Язык: Английский
Процитировано
3
Lecture notes in computer science, Год журнала: 2025, Номер unknown, С. 185 - 189
Опубликована: Янв. 1, 2025
Язык: Английский
Процитировано
0
Applied Mathematical Modelling, Год журнала: 2025, Номер 142, С. 115965 - 115965
Опубликована: Фев. 5, 2025
Язык: Английский
Процитировано
0
European Journal of Operational Research, Год журнала: 2025, Номер unknown
Опубликована: Фев. 1, 2025
Язык: Английский
Процитировано
0
Swarm and Evolutionary Computation, Год журнала: 2025, Номер 94, С. 101895 - 101895
Опубликована: Март 4, 2025
Язык: Английский
Процитировано
0
Australian & New Zealand Journal of Statistics, Год журнала: 2025, Номер unknown
Опубликована: Март 5, 2025
Summary In their paper ‘An Automatic Method for Solving Discrete Programming Problems’, Ailsa Land and Alison Doig developed a branch‐and‐bound method solving the general case of mixed integer linear programming (MIP) problem. A core part algorithm, branch variable selection, has received renewed attention in recent years with application machine learning methods to train new selection rules. this paper, we consider sources test instances used both these assess performance against existing methods. We apply instance space analysis (ISA) sufficiency generated cases purpose show how diversity can be intentionally increased support insights into many factors influencing MIP solver performance. The presents study comparing pseudocost branching full strong branching. propose ensure improved feature spaces that have been evolved more discriminating between different strategies are necessary add sufficient meaningful conclusions. While is small‐scale illustration need diverse instances, proposed approach generalisable tackle future exploration
Язык: Английский
Процитировано
0