Pub. Date | : Dec, 2019 |
---|---|
Product Name | : The IUP Journal of Information Technology |
Product Type | : Article |
Product Code | : IJIT21912 |
Author Name | : Bijaya Kumar Nanda and Satchidananda Dehuri |
Availability | : YES |
Subject/Domain | : Engineering |
Download Format | : PDF Format |
No. of Pages | : 18 |
The paper presents a K-mediod clustering algorithm based on ant colony optimization theory. Finding clusters in data is a challenging problem. Clustering task aims at the unsupervised classification of patterns in different clusters. Clustering with ant-based algorithm is the most promising method and has been shown to produce good results in a wide variety of problems. With this motivation, the paper proposes a novel ant-based K-mediods algorithm that modifies the ant K-means as locating the objects in a cluster with the probability, which is updated by the pheromone, while the rule of updating pheromone is according to the value, which is defined as the ratio of the total variation and within cluster variation.An experiment is conducted on both artificial and real-life databases. The experimental outcome confirms that clusters obtained through ant-based K-mediods are significantly better than clusters obtained through K-means and ant K-means in all databases.
Clustering is considered as an unsupervised classification process (Jain et al., 1999). Clustering is the process of grouping objects into clusters such that the objects from the same clusters are similar and objects from different clusters are dissimilar (Anderberg, 1973). Cluster analysis is also viewed as a tool for exploring the structure of data (Rasmussen, 1992). The relationship is often expressed as similarity or dissimilarity measurement and is calculated through distance measure. A clustering technique can be broadly divided into three main types: overlapping, in other words, nonexclusive; partitional; and hierarchical (Han and Kamber, 2006). The last two are related to each other in that a hierarchical clustering is a nested sequence of partitional clustering, each of which represents a hard partition of the dataset into different number of mutually disjoint subsets.
Ant colony optimization, Clustering, Ant-based clustering, K-means, K-mediods