International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 60 - Number 17 |
Year of Publication: 2012 |
Authors: Messaoud Hameurlaine, Abdelouahab Moussaoui, Hadda Cherroun |
10.5120/9782-4314 |
Messaoud Hameurlaine, Abdelouahab Moussaoui, Hadda Cherroun . AntMeans: A New Hybrid Algorithm based on Ant Colonies for Complex Data Mining. International Journal of Computer Applications. 60, 17 ( December 2012), 6-12. DOI=10.5120/9782-4314
The artificial learning is a scientific discipline that covers many aspects of study in mathematics, statistics and algorithms. The techniques of artificial learning have been widely used in extracting knowledge from data, but we are now facing new challenges "complex data. " The improvement of data mining techniques has become demanded. After a synthetic study of the different techniques of artificial learning and meta heuristics, this paper propose a hybrid solution "AntMeans" based on tools from data mining: kmeans and AntClust (algorithm based on artificial ants) then this solution is applied firstly in spatial data mining (the extraction of areas in a map of forest fires), and secondly in image data mining (characterization of image regions).