Loading…
Kids Learning Optimizer: social evolution and cognitive learning-based optimization algorithm
This paper proposes a novel social cognitive learning-based metaheuristic called kids Learning Optimizer (KLO), inspired by the early social learning behavior of kids organized as families in societal setup. In a society, people are organized as family groups (parents and children) where they intera...
Saved in:
Published in: | Neural computing & applications 2024-10, Vol.36 (28), p.17417-17465 |
---|---|
Main Authors: | , , |
Format: | Article |
Language: | English |
Subjects: | |
Citations: | Items that this one cites |
Online Access: | Get full text |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | This paper proposes a novel social cognitive learning-based metaheuristic called kids Learning Optimizer (KLO), inspired by the early social learning behavior of kids organized as families in societal setup. In a society, people are organized as family groups (parents and children) where they interact with each other within and outside their family. This interaction plays a vital role in early brain development, mannerisms, and behavioral learning. Early childhood learning is affected by various stimuli in their surroundings at different stages of life (newborn, infant, toddler, pre-school). This idea motivated us to map the decentralized learning concept of families and their interactions into a new algorithm where search agents (individuals) are arranged/organized in families, and they interact with each other at different stages of life to find the optimal solution. The algorithm is tested against 116 challenging benchmark functions including 31 unimodal, 63 multimodal, 14 CEC'2017 functions, and eight constrained functions. The algorithm is compared with 10 state-of-the-art algorithms. Friedman’s Mean Rank (FMR) and Wilcoxon rank-sum test (WRS) are used to measure the performance of competing algorithms. In the first two experiments, unimodal and multimodal benchmark functions are used to measure the explorative and exploitative ability of the algorithm. FMR and WRS show KLO outperformed all other algorithms. In the third experiment, the proposed method is tested against 14 CEC'2017 functions and eight constrained functions along with three real life mechanical engineering optimization problems. KLO proved to be better or equivalent to other competing algorithms. |
---|---|
ISSN: | 0941-0643 1433-3058 |
DOI: | 10.1007/s00521-024-10009-4 |