A Comparative Analysis of Different Categorical Data Clustering Ensemble Methods in Data Mining

Abstract

Over the past decades, a prevalent amount of work has been done in the data clustering research under the unsupervised learning technique in Data mining. Moreover a myriad of algorithms and methods has been proposed focusing on clustering different data types, representation of cluster models, and accuracy rates of the clusters. However no single clustering algorithm proves to be the most efficient in providing best results. Accordingly in order to find the solution to this issue a new technique, called Cluster ensemble method was bloomed. This cluster ensemble is a good alternative approach for facing the cluster analysis problem. The main aspire of the cluster ensemble is to combine different clustering solutions in such a way to achieve accuracy and to improve the quality of individual data clustering. Due to the substantial and unremitting development of the new methods in the sphere of data mining, it is obligatory to make a critical analysis of the existing techniques and the future novelty. This paper reveals the comparative study of different cluster ensemble methods along with their features, systematic working process and the average accuracy and error rates of each ensemble methods. Consequently this theoretical and comprehensive analysis will be very useful for the community of clustering practitioners and also helps in deciding the most suitable one to rectify the problem in hand.

Ensemble cluster pruning via convex‐concave programming

“…The solution given by the clustering algorithm, for its part, differs with respect to the similarity criterion either within objects in the cluster or between clusters. Cluster validity indexes are one of the measures that enable comparison of the quality of clusters . Huang et al proposed an ensemble clustering method according to an ensemble‐driven cluster uncertainty estimation and local weighting approach.…”

“…There is no restriction on the generation of different clustering solutions in the first step. This step can be composed of (i) several clustering algorithms or different clustering models created by various parameters of the same clustering algorithm, (ii) several object representations, and (iii) the selection of different subspaces of the data or finding a projection of data …”

“…Cluster validity indexes are one of the measures that enable comparison of the quality of clusters. 34 Huang et al proposed an ensemble clustering method according to an ensemble-driven cluster uncertainty estimation and local weighting approach. The uncertainty of the clusters was evaluated by taking the cluster labels in the whole ensemble with respect to an entropic criterion, according to which they developed a new ensemble-based cluster validity index.…”

Cluster Ensembling - A Technical Review

“…The advantage is that we combine the different clustering datasets with different algorithms. [13][14]This Squared Error Adjacent Matrix algorithm is mainly based upon the similarity matrix which is defined as the co-association matrix. It has the high possibility of finding the final data partition without previously knowing the number of clusters or any value of the thresholds when similarity matrix is given.…”

Using Frequent Closed Pattern Mining to Solve a Consensus Clustering Problem

“…More details in the surveys by Ghaemi et al [8], Sarumathi et al [16], and Vega-Pons & Ruiz-Shulcloper [21].…”