Partitional Data Clustering Research Articles

Clustering with Minimum Spanning Trees: How Good Can It Be?

AbstractMinimum spanning trees (MSTs) provide a convenient representation of datasets in numerous pattern recognition activities. Moreover, they are relatively fast to compute. In this paper, we quantify the extent to which they are meaningful in low-dimensional partitional data clustering tasks. By identifying the upper bounds for the agreement between the best (oracle) algorithm and the expert labels from a large battery of benchmark data, we discover that MST methods can be very competitive. Next, we review, study, extend, and generalise a few existing, state-of-the-art MST-based partitioning schemes. This leads to some new noteworthy approaches. Overall, the Genie and the information-theoretic methods often outperform the non-MST algorithms such as K-means, Gaussian mixtures, spectral clustering, Birch, density-based, and classical hierarchical agglomerative procedures. Nevertheless, we identify that there is still some room for improvement, and thus the development of novel algorithms is encouraged.

Data clustering using leaders and followers optimization and differential evolution

Data clustering is an important research topic in data mining. Although cluster analysis based on optimization algorithms has attracted great attention, optimization-based techniques face difficulties due to the non-linear objective function and complicated search space. Leaders and followers optimization (LaF) introduced in the 2015 IEEE Congress on Evolutionary Computation, and differential evolution algorithm (DE) are two efficient evolutionary computation methods, and they own some special advantages. The key power of LaF is the exploration in multi-modal search spaces, but it has a poor performance in the exploitation. On the other hand, DE based on the DE/best/1 mutation operator significantly promotes the exploitation process. In this study, the strong properties of LaF and DE are combined to balance exploration and exploitation in the search space to discover the cluster centroids. Besides, the proposed clustering method, i.e., LaF-DE, does not need parameter settings, unlike the existing optimization-based partitional clustering methods. Hence, this study proposes a straightforward, parameter-free, and efficient novel hybrid algorithm for the optimization-based partitional data clustering problem. Many experiments on the functions from CEC2017 test suite show that LaF-DE has better optimization performance and higher stability than the state-of-the-art metaheuristic algorithms. LaF-DE has been compared with well-known clustering techniques on the UCI and Shape datasets. The experimental results and statistical tests indicate that LaF-DE outperforms the well-known partitional clustering methods on 8 of 12 datasets in terms of internal performance metrics. Besides, LaF-DE performs better than density peaks clustering on 9 of 12 datasets in terms of external performance metrics.

A Differential Evolution Algorithm With Adaptive Niching and K-Means Operation for Data Clustering.

Clustering, as an important part of data mining, is inherently a challenging problem. This article proposes a differential evolution algorithm with adaptive niching and k -means operation (denoted as DE_ANS_AKO) for partitional data clustering. Within the proposed algorithm, an adaptive niching scheme, which can dynamically adjust the size of each niche in the population, is devised and integrated to prevent premature convergence of evolutionary search, thus appropriately searching the space to identify the optimal or near-optimal solution. Furthermore, to improve the search efficiency, an adaptive k -means operation has been designed and employed at the niche level of population. The performance of the proposed algorithm has been evaluated on synthetic as well as real datasets and compared with related methods. The experimental results reveal that the proposed algorithm is able to reliably and efficiently deliver high quality clustering solutions and generally outperforms related methods implemented for comparisons.

A survey of nature-inspired algorithm for partitional data clustering

The aim of the clustering is representing the huge amount of data objects by a smaller number of clusters or groups based on similarity. It is a task of good data analysis tool that required a rapid and precise partitioning the vast amount of data sets. The clustering problem is bring simplicity in modelling data and plays major role in the process of data mining and knowledge discovery. In the early stage, there are many conventional algorithm are used to solve the problem of data clustering. But, those conventional algorithms do not meet the requirement of clustering problem. Hence, the nature-inspired based approaches have been applied to fulfil the requirements data clustering problem and it can manage the shortcoming of conventional data clustering algorithm. This present paper is conducting a comprehensive review about the data clustering problem, discussed some of the machine learning datasets and performance metrics. This survey paper can helps to researcher in to the next steps in future.

An efficient hybrid clustering method based on improved cuckoo optimization and modified particle swarm optimization algorithms

Partitional data clustering with K-means algorithm is the dividing of objects into smaller and disjoint groups that has the most similarity with objects in a group and most dissimilarity from the objects of other groups. Several techniques have been proposed to avoid the major limitations of K-Means such as sensitive to initialization and easily convergence to local optima. An alternative to solve the drawback of the sensitive to centroids’ initialization in K-Means is the K-Harmonic Means (KHM) clustering algorithm. However, KHM is sensitive to the noise and easily runs into local optima. Over the past decade, many algorithms are developed for solving this problems based on evolutionary method. However, each algorithm has its own advantages, limitations and shortcomings. In this paper, we combined K-Harmonic Means (KHM) clustering algorithm with an improved Cuckoo Search (ICS) and particle swarm optimization (PSO). ICS is intended to global optimum solution using Lévy flight method through changing radius in a dynamic and shrewd manner. Therefore, it is faster than standard cuckoo search. ICS is effected with PSO to avoid falling into local optima. The proposed algorithm, called ICMPKHM, solves the local optima problem of KHM with significant improvement on efficacy and stability. Experiments with benchmark datasets show that the proposed algorithm is quite insensitive to the centroids’ initialization. Comparative studies with other algorithms reveal that the proposed algorithm produce high quality and stable clustering results.

A Kalman filtering induced heuristic optimization based partitional data clustering

Clustering algorithms have regained momentum with recent popularity of data mining and knowledge discovery approaches. To obtain good clustering in reasonable amount of time, various meta-heuristic approaches and their hybridization, sometimes with K-Means technique, have been employed. A Kalman Filtering based heuristic approach called Heuristic Kalman Algorithm (HKA) has been proposed a few years ago, which may be used for optimizing an objective function in data/feature space. In this paper at first HKA is employed in partitional data clustering. Then an improved approach named HKA-K is proposed, which combines the benefits of global exploration of HKA and the fast convergence of K-Means method. Implemented and tested on several datasets from UCI machine learning repository, the results obtained by HKA-K were compared with other hybrid meta-heuristic clustering approaches. It is shown that HKA-K is atleast as good as and often better than the other compared algorithms.

Comprehensive Study and Analysis of Partitional Data Clustering Techniques

Data clustering has found significant applications in various domains like bioinformatics, medical data, imaging, marketing study and crime analysis. There are several types of data clustering such as partitional, hierarchical, spectral, density-based, mixture-modeling to name a few. Among these, partitional clustering is well suited for most of the applications due to the less computational requirement. An analysis of various literatures available on partitional clustering will not only provide good knowledge, but will also lead to find the recent problems in partitional clustering domain. Accordingly, it is planned to do a comprehensive study with the literature of partitional data clustering techniques. In this paper, thirty three research articles have been taken for survey from the standard publishers from 2005 to 2013 under two different aspects namely the technical aspect and the application aspect. The technical aspect is further classified based on partitional clustering, constraint-based partitional clustering and evolutionary programming-based clustering techniques. Furthermore, an analysis is carried out, to find out the importance of the different approaches that can be adopted, so that any new development in partitional data clustering can be made easier to be carried out by researchers.