Research on an Ant Colony ISODATA Algorithm for Clustering Analysis in Real Time Computer Simulation

  • Authors:

  • Ying Wang;Ren-Wang Li;Bin Li;Peng-Ju Zhang;Yao-Hui Li

  • Affiliations:

  • -;-;-;-;-

  • Venue:
  • DMAMH '07 Proceedings of the Second Workshop on Digital Media and its Application in Museum & Heritage
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper intends to propose an advanced clustering method, ant colony ISODATA algorithm (ACIA) in real time computer simulation. Ant colony algorithm is used as the method of cursory clustering based on ants piling up their corpses and classifying their young ones. ISODATA algorithm is applied to meticulous clustering.This algorithm has been implemented and tested on several simulated data sets. At the same time, the performance efficiency of ACIA is analyzed based on four parameters:intracluster dissimilarity degree, intercluster dissimilarity degree, misclassification rate and CPU performance time. The computational results show that it is better than three other algorithms: ant colony K-means algorithm (ACKA), ant colony genetic algorithm (ACGA) and genetic K-means algorithm (GKA). Keywords: ACIA, Ant Colony Algorithm, ISODATA , Clustering Analysis, Real Time Computer Simulation