A Survey of Clustering Algorithms for Big Data: Taxonomy and Empirical Analysis

Abstract

Clustering algorithms have emerged as an alternative powerful meta-learning tool to accurately analyze the massive volume of data generated by modern applications. In particular, their main goal is to categorize data into clusters such that objects are grouped in the same cluster when they are similar according to specific metrics. There is a vast body of knowledge in the area of clustering and there has been attempts to analyze and categorize them for a larger number of applications. However, one of the major issues in using clustering algorithms for big data that causes confusion amongst practitioners is the lack of consensus in the definition of their properties as well as a lack of formal categorization. With the intention of alleviating these problems, this paper introduces concepts and algorithms related to clustering, a concise survey of existing (clustering) algorithms as well as providing a comparison, both from a theoretical and an empirical perspective. From a theoretical perspective, we developed a categorizing framework based on the main properties pointed out in previous studies. Empirically, we conducted extensive experiments where we compared the most representative algorithm from each of the categories using a large number of real (big) data sets. The effectiveness of the candidate clustering algorithms is measured through a number of internal and external validity metrics, stability, runtime, and scalability tests. In addition, we highlighted the set of clustering algorithms that are the best performing for big data.INDEX TERMS Clustering algorithms, unsupervised learning, big data.

New Pattern Reduction Method for Generalized Regression Neural Network

“…In density-based clustering, clusters are denoted with the density of the regions. The cluster grows in any direction leading by its density and each cluster has to contain at least minimum number of data in a neighborhood of a given radius [3]. In grid based algorithms, initially clustering space is divided into finite number of cells.…”

“…Then, clustering is performed on the grid [6]. Model based approaches begin with random parameter sets and they adjust these parameter sets over the iterations to find maximum likelihood estimator [3]. Similar to other optimization problems, clustering is also a kind of NP-hard problem and one of the biggest problem encountered in clustering approaches is to estimate the number of clusters.…”

Analysis of Hierarchical Clustering Algorithm to Handle Large Dataset

“…However, such techniques cannot correct mistaken decisions that once have taken. There are two approaches that can help in improving the quality of hierarchical clustering: (1) Firstly to perform carefu l analysis of object lin kages at each hierarchica l partit ioning or (2) By integrating hierarchical agglo merat ion and other approaches by first using a hierarchical agglo merative algorith m to group objects into micro-clusters, and then performing macro-clustering on the micro-clusters using another clustering method such as iterative relocation. One pro mising direct ion for improving the clustering quality of h ierarchical methods is to integrate hierarchical clustering with other clustering techniques for mu ltiple phase clustering.…”

“…One pro mising direct ion for improving the clustering quality of h ierarchical methods is to integrate hierarchical clustering with other clustering techniques for mu ltiple phase clustering. So in order to make improvement in hierarchical clustering we merge some other techniques or method in to it [1]. Every day, 2.5 quintillion bytes of data are created and 90 percent o f the data in the world today were produced within the past two years.…”

Comprehensive Analysis & Performance Comparison of Clustering Algorithms for Big Data

“…Partitioning-Based Clustering Algorithms [9] In partitioning-based algorithms, the data is distributed into various data subsets. The reason behind this splitting is lack of feasibility to check every possible subset; there are certain greedy probing schemes which are used in form of iterative inflation.…”