A Data Clustering Approach: Using Ant Colony Optimization

Ant colony optimization has been successfully applied to many combinatorial optimization problem. It has been firstly implemented for traveling salesman problem which is well known problem in engineering science. One chemical substance called pheromone plays very important role in any ant colony optimization algorithm. Apart from foraging behavior another behavior called brood shorting is also notify in some species of real ants.Ants will classify to that cluster which is having more similar to that ant and like that the whole tree is built. The branches to the root is considered to be independent clusters discovered by this algorithm and by this way the number of clusters identified by this approach is the number of braches made to the root node of the AntTree. Here depending on order of we feed the ant as input to the algorithm sometimes it might be possible to wrongly cluster the data. Thus it will increase the probability of clustering algorithm.