Partition Based Clustering Algorithms Research Articles

Energy aware clustered load balancing in cloud computing environment

Cloud is a collection of datacentres with heterogeneous resources, which gives services to the users based on pay-as-you-use model. Even though it has several advantages such as availability, scalability, and reliability, some performance parameters like energy consumption, load balancing, response time, resource allocation time, etc., are not properly fine tuned. This paper proposes an energy aware clustered load balancing system in which, heterogeneous resources are clustered into different groups by using a partitioning-based clustering algorithm. The clustering reduces number of resources needs to be searched and hence minimizes the time required for resource discovery. An energy aware best-fit virtual machine (VM) allocation is used for reducing the power consumption. The process allocations to VMs are done based on best-fit allocation strategy for optimal space utilisation. The results show that proposed method reduces time for resource discovery, resource allocation and response time with power consumption.

A P System for K-Medoids-Based Clustering

The membrane computing model, also known as the P system, is a parallel and distributed computing system. K-medoids algorithm is one of the most famous algorithms in partition-based clustering algorithms, and has been widely used in data analysis and modern scientific research. Combining the P system with the k-medoids algorithm, the maximum parallelism calculated by the P system can effectively reduce the time complexity of the k-medoids clustering algorithm. Based on this, this paper proposes a cell-like P system with promoters and inhibitors based on k-medoids clustering, and then an instance is given to illustrate the practicability and effectiveness of the P system designed.

Survey on Partition based Clustering Algorithms in Big Data

Survey on Partition based Clustering Algorithms in Big Data

A State of Art Analysis of Telecommunication Data by k-Means and k-Medoids Clustering Algorithms

Cluster analysis is one of the major data analysis methods widely used for many practical applications in emerging areas of data mining. A good clustering method will produce high quality clusters with high intra-cluster similarity and low inter-cluster similarity. Clustering techniques are applied in different domains to predict future trends of available data and its uses for the real world. This research work is carried out to find the performance of two of the most delegated, partition based clustering algorithms namely k-Means and k-Medoids. A state of art analysis of these two algorithms is implemented and performance is analyzed based on their clustering result quality by means of its execution time and other components. Telecommunication data is the source data for this analysis. The connection oriented broadband data is given as input to find the clustering quality of the algorithms. Distance between the server locations and their connection is considered for clustering. Execution time for each algorithm is analyzed and the results are compared with one another. Results found in comparison study are satisfactory for the chosen application.

An Extended K-Means Algorithm using MapReduce Framework for Mixed Datasets

K-Means is a famous partition based clustering algorithm. Various extensions of KMeans have been proposed depending on the type of datasets being handled. Popular ones include K-Modes for categorical data and K-Prototype for mixed numerical and categorical data. The K-Means and its extensions suffer from one major limitation that is dependency on prior input of number of clusters K. Sometimes it becomes practically impossible to correctly estimate the optimum number of clusters in advance. Various ways have been suggested in literature to overcome this limitation for numerical data. But for categorical and mixed data work is still in progress. In this paper, we introduce a new algorithm based on the K-Means that takes mixed dataset as an input and generates appropriate number of clusters on the run using MapReduce programming style. The new algorithm not only overcomes the limitation of providing the value of K initially but also reduces the computation time using MapReduce framework.

An evaluation-committee recommendation system for national R&D projects using social network analysis

Korean National Science & Technology Information Service (NTIS) provides a service of evaluating national R&D projects and providing such evaluated national R&D projects along with their participating researcher information. It also provides a service of recommending and selecting evaluation committees for the R&D projects. Transparency is an important aspect that should be ensured on the evaluation process of the national R&D projects. Thus, the recommending unfamiliar evaluation committees with the participants of the R&D projects are one of the important aspects that can ensure the transparency for the evaluation process. In this paper, we present an evaluation-committee recommendation system using an online detection method of researcher connections by a partitioning-based clustering algorithm and random walks. The clustering algorithm enables us to partition the network to number of small graphs that can be processed via random walks. Then, we can rank the connection weight of each suspicious researcher according to a researcher in charge of a R&D project and we can exclude the researchers having higher connection weight from the evaluation committee of the R&D project. In addition, we also present a text-data refinement and entity identification method using Jaro---Winkler distance algorithm to construct more precise researcher network.

Performance comparison of fuzzy and non-fuzzy classification methods

In data clustering, partition based clustering algorithms are widely used clustering algorithms. Among various partition algorithms, fuzzy algorithms, Fuzzy c-Means (FCM), Gustafson–Kessel (GK) and non-fuzzy algorithm, k-means (KM) are most popular methods. k-means and Fuzzy c-Means use standard Euclidian distance measure and Gustafson–Kessel uses fuzzy covariance matrix in their distance metrics. In this work, a comparative study of these algorithms with different famous real world data sets, liver disorder and wine from the UCI repository is presented. The performance of the three algorithms is analyzed based on the clustering output criteria. The results were compared with the results obtained from the repository. The results showed that Gustafson–Kessel produces close results to Fuzzy c-Means. Further, the experimental results demonstrate that k-means outperforms the Fuzzy c-Means and Gustafson–Kessel algorithms. Thus the efficiency of k-means is better than that of Fuzzy c-Means and Gustafson–Kessel algorithms.

Clustering and Cooperative Beamforming in Green Cognitive Radio Networks with Relays

A relay assisted cooperative beamforming technique can efficiently improve the spectral efficiency and can reduce the interference to primary users in a cognitive underlay systems. Our objective is to reduce the total power cost of Secondary Users (SU) while limiting the interference in the direction of Primary Users (PU). Kmean algorithm which is a partition based clustering algorithm is used here as the clustering algorithm. While considering flexible cooperation as general scheme, we collectively optimize the clustering and the beamforming to lower the overall power consumption while fulfilling the secondary users quality of service and satisfying the primary users interference temperature limits. An iterative algorithm called generalized Benders decomposition method is used to find an optimal solution. The indirect transmission mode of relays is considered in this work. The relays are equipped with directional antennas which will reduce the interference to the PU and meet the power cost requirement of SU under the assumptions that the clusters are non-overlapping and fixed. Thus spectral efficiency can be increased. Through simulations, we analyzed the advantages of algorithms and showed the benefits of the relays in cognitive wireless networks.

Observer-Biased Fuzzy Clustering

As generated by clustering algorithms, clusterings (or partitions) are hypotheses on data explanation which are better evaluated by experts from the application domain. In general, clustering algorithms allow a limited usage of domain knowledge about the cluster formation process. In this study, we propose both a design technique and a new partitioning-based clustering algorithm which can be used to assist the data analyst while looking for a set of meaningful clusters, i.e., clusters that actually correspond to the underlying data structure. Following an observer metaphor according to which the perception of a group of objects depends on the observer position-the closer an observer is from an image more details (s)he perceives-we resort to shrinkage to incorporate a regularization term, accounting for the observation point, within the objective function of an otherwise unbiased clustering algorithm. This technique allows our resulting biased algorithm to generate a set of reasonable partitions, i.e., partitions validated by a given cluster validity index, corresponding to views of data with different levels of granularity (levels of detail) in different regions of the data space. For the illustration of the design technique, we adopted the fuzzy c-means (FCM) algorithm as the unbiased clustering algorithm and include a convergence theorem assuring that changing the point of observation in the corresponding biased algorithm FCM with focal point (FCMFP) does not jeopardize its convergence. Experimental studies on both synthetic and real data are included to illustrate the usefulness of the approach. In addition, and as a convenient side effect of using shrinkage, the experimental results suggest that our biased algorithm (FCMFP) not only seems to scale better than the successive runs of the unbiased one (FCM) but on the average, seems to produce clusters exhibiting higher validity index values as well. In addition, less sensitivity to initialization was observed for the biased algorithm when compared with the unbiased one.

English

K-means is the well known and most familiar algorithm among the other partition based clustering algorithms. It typically shows spectacular results even in significantly massive information sets of Image segmentation supported associate adaptive K-means clustering algorithm is conferred. The proposed method tries to develop Kmeans algorithm to get high performance and potency. This technique proposes data formatting step in K-means algorithmic rule. additionally, it solves a model choice variety by deciding the quantity of clusters victimization datasets from image by frame size and also the definite quantity between the means that, and extra steps for convergence step in K-means algorithm are supplementary. Moreover, so as to judge the performance of the proposed technique, the results of the proposed technique, customary K-means and recently changed K-means are compared. The experimental results showed that the proposed technique provides higher output.

Improving the Initial Centroids of k-means Clustering Algorithm to Generalize its Applicability

k-means is one of the most widely used partition based clustering algorithm. But the initial centroids generated randomly by the k-means algorithm cause the algorithm to converge at the local optimum. So to make k-means algorithm globally optimum, the initial centroids must be selected carefully rather than randomly. Though many researchers have already been carried out for the enhancement of k-means algorithm, they have their own limitations. In this paper a new method to formulate the initial centroids is proposed which results in better clusters equally for uniform and non-uniform data sets.

Performance based analysis between k-Means and Fuzzy C-Means clustering algorithms for connection oriented telecommunication data

Data mining is the process of discovering meaningful new correlation, patterns and trends by sifting through large amounts of data, using pattern recognition technologies as well as statistical and mathematical techniques. Cluster analysis is often used as one of the major data analysis technique widely applied for many practical applications in emerging areas of data mining. Two of the most delegated, partition based clustering algorithms namely k-Means and Fuzzy C-Means are analyzed in this research work. These algorithms are implemented by means of practical approach to analyze its performance, based on their computational time. The telecommunication data is the source data for this analysis. The connection oriented broad band data is used to find the performance of the chosen algorithms. The distance (Euclidian distance) between the server locations and their connections are rearranged after processing the data. The computational complexity (execution time) of each algorithm is analyzed and the results are compared with one another. By comparing the result of this practical approach, it was found that the results obtained are more accurate, easy to understand and above all the time taken to process the data was substantially high in Fuzzy C-Means algorithm than the k-Means.

PFClust: an optimised implementation of a parameter-free clustering algorithm

BackgroundA well-known problem in cluster analysis is finding an optimal number of clusters reflecting the inherent structure of the data. PFClust is a partitioning-based clustering algorithm capable, unlike many widely-used clustering algorithms, of automatically proposing an optimal number of clusters for the data.ResultsThe results of tests on various types of data showed that PFClust can discover clusters of arbitrary shapes, sizes and densities. The previous implementation of the algorithm had already been successfully used to cluster large macromolecular structures and small druglike compounds. We have greatly improved the algorithm by a more efficient implementation, which enables PFClust to process large data sets acceptably fast.ConclusionsIn this paper we present a new optimized implementation of the PFClust algorithm that runs considerably faster than the original.

Efficient stochastic algorithms for document clustering

Clustering has become an increasingly important and highly complicated research area for targeting useful and relevant information in modern application domains such as the World Wide Web. Recent studies have shown that the most commonly used partitioning-based clustering algorithm, the K-means algorithm, is more suitable for large datasets. However, the K-means algorithm may generate a local optimal clustering. In this paper, we present novel document clustering algorithms based on the Harmony Search (HS) optimization method. By modeling clustering as an optimization problem, we first propose a pure HS based clustering algorithm that finds near-optimal clusters within a reasonable time. Then, harmony clustering is integrated with the K-means algorithm in three ways to achieve better clustering by combining the explorative power of HS with the refining power of the K-means. Contrary to the localized searching property of K-means algorithm, the proposed algorithms perform a globalized search in the entire solution space. Additionally, the proposed algorithms improve K-means by making it less dependent on the initial parameters such as randomly chosen initial cluster centers, therefore, making it more stable. The behavior of the proposed algorithm is theoretically analyzed by modeling its population variance as a Markov chain. We also conduct an empirical study to determine the impacts of various parameters on the quality of clusters and convergence behavior of the algorithms. In the experiments, we apply the proposed algorithms along with K-means and a Genetic Algorithm (GA) based clustering algorithm on five different document datasets. Experimental results reveal that the proposed algorithms can find better clusters and the quality of clusters is comparable based on F-measure, Entropy, Purity, and Average Distance of Documents to the Cluster Centroid (ADDC).

A Survey of Partition based Clustering Algorithms in Data Mining: An Experimental Approach

A Survey of Partition based Clustering Algorithms in Data Mining: An Experimental Approach

Cluster analysis on time series gene expression data

Cluster analysis is frequently used to study the trend of gene expression behaviours from microarray time series data. We adopt a partitioning-based clustering algorithm for such a task. After time series are discritised into sequences, a sequential pattern mining technique is applied to find patterns as the initial clusters. Longest Common Subseries Similarity is used to measure the similarity between time series which overcomes the 'shift-effect' influence. An object is re-assigned to the cluster which has most objects within the k nearest neighbours of the object. Similarity measurements, like Pearson correlation coefficient, are used to determine the neighbours.

Effective hybrid PSO and K-means clustering algorithm for gene expression data

DNA micro-array technology helps monitor the expression levels of thousands of genes. This paper presents clustering of gene expression data using particle swarm optimisation (PSO) and K-means algorithm combined. Recent studies have shown that partition-based clustering algorithms are more suitable for clustering large datasets. Partition-based K-means is used mostly because of its simple implementation and fast convergence. But it suffers local optima. PSO is a population-based stochastic search process, which searches automatically for the optimal solution in the search space. So, it is combined with K-means algorithm for clustering. In the previous PSO employing papers, particles flock at boundary. Our technique removes boundary blocks by moving the boundary particles towards the global best particle to improve effectiveness of PSO. The results of hybrid PSO, K-means and PSO algorithms are compared for several datasets. Among the three algorithms, the hybrid PSO algorithm performs well for most of the ...

Harmony K-means algorithm for document clustering

Fast and high quality document clustering is a crucial task in organizing information, search engine results, enhancing web crawling, and information retrieval or filtering. Recent studies have shown that the most commonly used partition-based clustering algorithm, the K-means algorithm, is more suitable for large datasets. However, the K-means algorithm can generate a local optimal solution. In this paper we propose a novel Harmony K-means Algorithm (HKA) that deals with document clustering based on Harmony Search (HS) optimization method. It is proved by means of finite Markov chain theory that the HKA converges to the global optimum. To demonstrate the effectiveness and speed of HKA, we have applied HKA algorithms on some standard datasets. We also compare the HKA with other meta-heuristic and model-based document clustering approaches. Experimental results reveal that the HKA algorithm converges to the best known optimum faster than other methods and the quality of clusters are comparable.

GRINDING WHEEL CONDITION MONITORING WITH HIDDEN MARKOV MODEL-BASED CLUSTERING METHODS

Hidden Markov model (HMM) is well known for sequence modeling and has been used for condition monitoring. However, HMM-based clustering methods are developed only recently. This article proposes a HMM-based clustering method for monitoring the condition of grinding wheel used in grinding operations. The proposed method first extract features from signals based on discrete wavelet decomposition using a moving window approach. It then generates a distance (dissimilarity) matrix using HMM. Based on this distance matrix several hierarchical and partitioning-based clustering algorithms are applied to obtain clustering results. The proposed methodology was tested with feature sequences extracted from acoustic emission signals. The results show that clustering accuracy is dependent upon cutting condition. Higher material removal rate seems to produce more discriminatory signals/features than lower material removal rate. The effect of window size, wavelet decomposition level, wavelet basis, clustering algorithm, and data normalization were also studied.

Empirical comparison of fast partitioning-based clustering algorithms for large data sets

Several fast algorithms for clustering very large data sets have been proposed in the literature, including CLARA, CLARANS, GAC-R3, and GAC-RARw. CLARA is a combination of a sampling procedure and the classical PAM algorithm, while CLARANS adopts a serial randomized search strategy to find the optimal set of medoids. GAC-R3 and GAC-RARw exploit genetic search heuristics for solving clustering problems. In this research, we conducted an empirical comparison of these four clustering algorithms over a wide range of data characteristics described by data size, number of clusters, cluster distinctness, cluster asymmetry, and data randomness. According to the experimental results, CLARANS outperforms its counterparts both in clustering quality and execution time when the number of clusters increases, clusters are more closely related, more asymmetric clusters are present, or more random objects exist in the data set. With a specific number of clusters, CLARA can efficiently achieve satisfactory clustering quality when the data size is larger, whereas GAC-R3 and GAC-RARw can achieve satisfactory clustering quality and efficiency when the data size is small, the number of clusters is small, and clusters are more distinct and symmetric.