TY - JOUR
T1 - Opposition-Inspired synergy in sub-colonies of ants
T2 - The case of Focused Ant Solver
AU - Rojas-Morales, Nicolás
AU - Riff, María Cristina
AU - Montero, Elizabeth
N1 - Publisher Copyright:
© 2021
PY - 2021/10/11
Y1 - 2021/10/11
N2 - Recently, Opposition-Inspired Learning strategies were proposed to improve the search process of ant-based algorithms to solve combinatorial problems. In this paper, we propose a collaborative framework of these strategies called Multiple Opposite Synergic Strategy for Ants (MOSSA). Because of each strategy has a different goal, we expect that the ants algorithm will benefit from their collaboration. The algorithm strongly uses the pheromone matrix for accomplishing stigmergy. To evaluate our framework, we use a recently proposed algorithm to solve Constraint Satisfaction Problems named Focused Ant Solver. Results and statistical analysis show that using MOSSA, Focused Ant Solver is able to solve more problems from the transition phase.
AB - Recently, Opposition-Inspired Learning strategies were proposed to improve the search process of ant-based algorithms to solve combinatorial problems. In this paper, we propose a collaborative framework of these strategies called Multiple Opposite Synergic Strategy for Ants (MOSSA). Because of each strategy has a different goal, we expect that the ants algorithm will benefit from their collaboration. The algorithm strongly uses the pheromone matrix for accomplishing stigmergy. To evaluate our framework, we use a recently proposed algorithm to solve Constraint Satisfaction Problems named Focused Ant Solver. Results and statistical analysis show that using MOSSA, Focused Ant Solver is able to solve more problems from the transition phase.
KW - Ant colony optimization
KW - Constraint Satisfaction Problems
KW - Opposition-Inspired Learning
UR - http://www.scopus.com/inward/record.url?scp=85111304485&partnerID=8YFLogxK
U2 - 10.1016/j.knosys.2021.107341
DO - 10.1016/j.knosys.2021.107341
M3 - Article
AN - SCOPUS:85111304485
SN - 0950-7051
VL - 229
JO - Knowledge-Based Systems
JF - Knowledge-Based Systems
M1 - 107341
ER -