Ant colony optimization is a successful metaheuristic for solving combinatorial optimization problems. However, the drawback of premature exploitation arises in ant colony optimization when coupled with local searches, in which the neighborhood's structures of the search space are not completely traversed. This paper proposes two algorithmic components for solving the premature exploitation, i.e. the reactive heuristics and recursive local search technique. The resulting algorithm is tested on two well-known combinatorial optimization problems arising in the artificial intelligence problems field and compared experimentally to six (6) variants of ACO with local search. Results showed that the enhanced algorithm outperforms the six ACO variants.
Ant colony optimization is one of the most successful examples of swarm intelligent systems. The exploration and exploitation are the main mechanisms in controlling search within the ACO. Reactive search is a framework for automating the exploration and exploitation in stochastic algorithms. Restarting the search with the aid of memorizing the search history is the soul of reaction. It is to increase the exploration only when needed. This paper proposes a reactive memory model to overcome the limitation of the random exploration after restart because of losing the previous history of search. The proposed model is utilized to record the previous search regions to be used as reference for ants after restart. The performances of six (6) ant colony optimization variants were evaluated to select the base for the proposed model. Based on the results, Max-Min Ant System has been chosen as the base for the modification. The modified algorithm called RMMAS, was applied to TSPLIB95 data and results showed that RMMAS outperformed the standard MMAS.
This study proposes machine learning strategies to control the parameter adaptation in ant colony optimization algorithm, the prominent swarm intelligence metaheuristic. The sensitivity to parameters' selection is one of the main limitations within the swarm intelligence algorithms when solving combinatorial problems. These parameters are often tuned manually by algorithm experts to a set that seems to work well for the problem under study, a standard set from the literature or using off-line parameter tuning procedures. In the present study, the parameter search process is integrated within the running of the ant colony optimization without incurring an undue computational overhead. The proposed strategies were based on a novel nature-inspired idea. The results for the travelling salesman and quadratic assignment problems revealed that the use of the augmented strategies generally performs well against other parameter adaptation methods.
A statistical machine learning indicator, ACOustic, is proposed to evaluate the exploration behavior in the iterations of ant colony optimization algorithms. This idea is inspired by the behavior of some parasites in their mimicry to the queens' acoustics of their ant hosts. The parasites' reaction results from their ability to indicate the state of penetration. The proposed indicator solves the problem of robustness that results from the difference of magnitudes in the distance's matrix, especially when combinatorial optimization problems with rugged fitness landscape are applied. The performance of the proposed indicator is evaluated against the existing indicators in six variants of ant colony optimization algorithms. Instances for travelling salesman problem and quadratic assignment problem are used in the experimental evaluation. The analytical results showed that the proposed indicator is more informative and more robust.
This intensification and diversification in AntColony Optimization (ACO) is the search strategy to achieve a trade-off between learning a new search experience (exploration) and earning from the previous experience (exploitation). The automation between the two processes is maintained using reactive search. However, existing works in ACO were limited either to the management of pheromone memory or to the adaptation of few parameters. This paper introduces the reactive ant colony optimization (RACO) strategy that sticks to the reactive way of automation using memory, diversity indication, and parameterization. The performance of RACO is evaluated on the travelling salesman and quadratic assignment problems from TSPLIB and QAPLIB, respectively. Results based on a comparison of relative percentage deviation revealed the superiority of RACO over other well-known metaheuristics algorithms. The output of this study can improve the quality of solutions as exemplified by RACO.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.