An overview of various enhancements of DENCLUE algorithm
Mariam S. Khader, Ghazi Al‐Naymat · 2019
Clustering is one of the main data mining methods for knowledge discovery. The clustering is an exploratory data analysis technique that categorizes different data objects into similar groups, named clusters. Density-based clustering defines clusters as dense regions that are separated by low dense regions. The DENCLUE (DENsity CLUstEring) is a robust density-based algorithm for discovering clusters with arbitrary shapes and sizes. Although its efficiency, the DENCLUE suffers from the following issues: (1) It is sensitive to the values of its parameters. (2) It fails to discover clusters with highly varying densities. (3) It may require large computation time for clustering large datasets. Several research articles attempted to enhance the performance of the DENCLUE algorithm to overcome the mentioned issues. This research surveys the proposed enhancements of the DENCLUE algorithm concerning their main contribution, input parameters, and evaluation measures. The research aims to serve as a base for future enhancements of the DENCLUE algorithm.