Hybrid data clustering approaches using bacterial colony optimization and k-means

J. Revathi, V. P. Eswaramurthy, P. Padmavathi

2021IOP Conference Series Materials Science and Engineering16 citationsDOI Open Access PDF

Abstract

Abstract Data clustering is a fashionable data analysis technique in the data mining. K-means is a popular clustering technique for solving a clustering problem. However, the k-means clustering technique extremely depends on the initial position and converges to a local optimum. On the other hand, the bacterial colony optimization (BCO) is a well-known recently proposed data clustering algorithm. However, it is a high computational cost to complete a given solution. Hence, this research paper proposes a new hybrid data clustering method for solving data clustering problem. The proposed hybrid data clustering algorithm is a combination of the BCO and K-means called BCO+KM clustering algorithm. The experimental result shows that the proposed hybrid BCO+KM data clustering algorithm reveal better cluster partitions.

Topics & Concepts

Cluster analysisCURE data clustering algorithmCorrelation clusteringCanopy clustering algorithmData miningComputer scienceData stream clusteringDetermining the number of clusters in a data setClustering high-dimensional dataSingle-linkage clusteringFuzzy clusteringAlgorithmArtificial intelligenceAdvanced Clustering Algorithms ResearchMetaheuristic Optimization Algorithms ResearchData Stream Mining Techniques