An analytical approach to the similarities between swarm intelligence and artificial neural network

Abstract

In this paper, the authors focus on swarm intelligence represented by the ant colony system and artificial neural network simulating the brain working principles, and analyse the similarities between them by the approaches of a complexity study. Firstly, the similarities between swarm intelligence and neural network have been analysed qualitatively from system structure and operation mechanisms and the results show that there exist obvious similarities in overall disciplines between them. Moreover, the mapping relationships between swarm intelligence and artificial neural network at algorithm level have been revealed using the travelling salesman problem (TSP), which illustrates that these two algorithms are also similar from the viewpoint of numeric computing and simulating when they are directly used to solve a complex problem. In addition, the swarm intelligence algorithm and the artificial neural network algorithm are adopted to solve an engineering practical problem, which can be transformed into the TSP. The results show that the performance of these two algorithms still has close similarities. The purpose of the above work is to construct the inherent relationship among different bio-inspired computation methods, which has both important theoretic significance and practical value to reveal the generation and operation mechanisms of human intelligence.

Complete receptor editing operation based on quantum clonal selection algorithm for optimization problems

“…HMPP aims to minimize the overall machining costs, including the moving cost of tools ( C M ), cutting tools cost ( C T ), and tools replacement cost ( C R ). The mathematical formulation of the case is as follows : …”

Klasifikasi Algoritma Swarm Intelligence Dalam Perspektif Complex Adaptive System dengan Metode Uji Komparasi Statistik

“…Xiao dkk [11] melakukan pendekatan analisis kemiripan antara SI dan jaringan syaraf tiruan (JST). Penelitian ini bertujuan untuk mengklasifikasikan SI berdasarkan kesamaan karakteristiknya terhadap CAS.…”

“…[1] yang di dalamnya mengungkapkan perlunya melihat perspektif klasifikasi algoritma SI, bio-inspired algorithm (BIA), dan natural inspired algorithm dari sudut pandang CAS. Selain itu, penelitian ini telah memberikan sisi lain pemetaan SI berdasarkan karakteristik CAS, seperti [10] yang memetakan berdasarkan mekanisme biomimetik dan [11] …”

Improved Cat Swarm Optimization Algorithm for Assembly Sequence Planning

“…Evolution algorithms provide new solutions to various complex optimization problems by imitating the self-*Address correspondence to this author at the School of Mechanical Engineering, Dongguan University of Technology, Dongguan 523808, China; Tel: +86 13532853596; E-mail: guojw@dgut.edu.cn organization mechanism of natural biological communities and the adaptability of evolution [7][8][9]. The evolution algorithms that have been investigated for ASP include the genetic algorithms [10], ant colony [11], and particle swarm optimization [12].…”