Distance-based Clustering Algorithm Research Articles

A new local density and relative distance based spectrum clustering

A novel local density and relative distance-based spectrum clustering (LDRDSC) algorithm is proposed for multidimensional data clustering. The density spectra consider both redefined local densities and relative distances. The spectral peaks are defined as cluster centers since these peaks correspond to the local density maximums. Different clusters correspond to different spectra. The clustering by fast search and find of density peaks (CFSFDP) algorithm and several benchmark data sets are employed to validate our proposed LDRDSC algorithm. Once the density spectrum is generated, the rest points can be automatically clustered by our LDRDSC algorithm, which is different from CFSFDP. CFSFDP needs to categorize data points according to the cluster centers. Furthermore, our LDRDSC algorithm is compared with other five typical clustering algorithms (DBSCAN, FCM, AP, Mean Shift and k-means) in order to validate the effectiveness of the proposed algorithm. Computational results demonstrate that our algorithm can obtain a better clustering result than the above mentioned algorithms, especially in identifying noises or isolates.

Unsupervised Classification of Multivariate Time Series Using VPCA and Fuzzy Clustering With Spatial Weighted Matrix Distance.

Due to high dimensionality and multiple variables, unsupervised classification of multivariate time series (MTS) involves more challenging problems than those of univariate ones. Unlike the vectorization of a feature matrix in traditional clustering algorithms, an unsupervised pattern recognition scheme based on matrix data is proposed for MTS samples in this paper. To reduce the computational load and time consumption, a novel variable-based principal component analysis (VPCA) is first devised for the dimensionality reduction of MTS samples. Afterward, a spatial weighted matrix distance-based fuzzy clustering (SWMDFC) algorithm is proposed to directly group MTS samples into clusters as well as preserve the structure of the data matrix. The spatial weighted matrix distance (SWMD) integrates the spatial dimensionality difference of elements of data into the distance of MST pairs. In terms of the SWMD, the MTS samples are clustered without vectorization in the dimensionality-reduced feature matrix space. Finally, three open-access datasets are utilized for the validation of the proposed unsupervised classification scheme. The results show that the VPCA can capture more features of MTS data than principal component analysis (PCA) and 2-D PCA. Furthermore, the clustering performance of SWMDFC is superior to that of fuzzy c -means clustering algorithms based on the Euclidean distance or image Euclidean distance.

Intra-Cluster Distance Minimization in DNA Methylation Analysis Using an Advanced Tabu-Based Iterative -Medoids Clustering Algorithm (T-CLUST)

Recent advances in DNA methylation profiling have paved the way for understanding the underlying epigenetic mechanisms of various diseases such as cancer. While conventional distance-based clustering algorithms (e.g., hierarchical and k-means clustering) have been heavily used in such profiling owing to their speed in conduct of high-throughput analysis, these methods commonly converge to suboptimal solutions and/or trivial clusters due to their greedy search nature. Hence, methodologies are needed to improve the quality of clusters formed by these algorithms without sacrificing from their speed. In this study, we introduce three related algorithms for a complete high-throughput methylation analysis: a variance-based dimension reduction algorithm to handle high-dimensionality in data, an outlier detection algorithm to identify the outliers of data, and an advanced Tabu-based iterative k-medoids clustering algorithm (T-CLUST) to reduce the impact of initial solutions on the performance of conventional k-medoids algorithm. The performance of the proposed algorithms is demonstrated on nine different real DNA methylation datasets obtained from the Gene Expression Omnibus DataSets database. The accuracy of the cluster identification obtained by our proposed algorithms is higher than those of hierarchical and k-means clustering, as well as the conventional methods. The algorithms are implemented in MATLAB, and available at: http://www.coe.miami.edu/simlab/tclust.html.

Energy proficient clustering technique for lifetime enhancement of cognitive radio–based heterogeneous wireless sensor network

Utilizing the available spectrum in a more optimized manner and selecting a proper routing technique for transferring the data, without any data collision, from the sensor node to the base station play a major role in any network for increasing their network lifetime. Cognitive radio techniques play a major role to achieve the same, and when combined with wireless sensor networks the above-said requirements can be greatly accomplished. In this article, a novel energy-efficient distance-based clustering and routing algorithm using multi-hop communication approach is proposed. Based on distance, the given heterogeneous cognitive radio–based wireless sensor networks are divided into regions and are allocated with a unique spectrum. Dynamic clustering through distance calculation and routing of data through multi-hop communication is done. The simulation results illustrate that the proposed algorithm has improved energy efficiency and is more stable. The first node death and 80% node death illustrate the improved scalability. Also, the increased throughput aids in maintaining the residual energy of the network, which further solves the problem of load balancing among nodes. All the above results combined with half node death analysis show that the proposed algorithm also has an improved network lifetime.

Genetic algorithm based fault tolerant clustering in wireless sensor network

In wireless sensor network (WSN), during fault detection and recovery, average energy loss, and message loss occurs including the link failure. Also, the number of faulty nodes and the traffic overhead is increased with the size of WSN. In order to overcome this issue, in this study, a cluster-based fault tolerance technique using genetic algorithm is proposed. Here the network is clustered according to energy-efficient distance-based clustering algorithm. For each cluster head, a set of backup nodes are selected using genetic algorithm based on the sponsored coverage and residual energy parameters. This helps in detecting the faults occurring in cluster members and cluster heads. Simulation results show that the proposed technique minimises the energy loss and overhead.

On Using a Clustering Approach for Global Climate Classification

Abstract Classifying the land surface into climate types provides means of diagnosing relations between Earth’s physical and biological systems and the climate. Global climate classifications are also used to visualize climate change. Clustering climate datasets provides a natural approach to climate classification, but the rule-based Köppen–Geiger classification (KGC) is the one most widely used. Here, a comprehensive approach to the clustering-based classification of climates is presented. Local climate is defined as a multivariate time series of mean monthly climatic variables and the authors propose to use dynamic time warping (DTW) as a measure of dissimilarity between local climates. Also discussed are the choice of climatic variables, the importance of their proper normalization, and the advantage of using distance-based clustering algorithms. Using the WorldClim global climate dataset and different combinations of clustering parameters, 32 different clustering-based classifications are calculated. These classifications are compared between themselves and to the KGC using the information-theoretic V measure. It is found that the best classifications are obtained using three climate variables (temperature, precipitation, and temperature range), a data normalization that takes into account the skewed distribution of precipitation values, and the partitioning around medoids clustering algorithm. Two such classifications are compared in detail between each other and to the KGC. About half of the climate types found by clustering can be matched to the familiar KGC classes, but the rest differ in their climatic character and spatial distribution. Finally, it is demonstrated that clustering-based classification results in climate types that are internally more homogeneous and externally more distinct than climate types in the KGC.

The Gap Procedure: for the identification of phylogenetic clusters in HIV-1 sequence data.

BackgroundIn the context of infectious disease, sequence clustering can be used to provide important insights into the dynamics of transmission. Cluster analysis is usually performed using a phylogenetic approach whereby clusters are assigned on the basis of sufficiently small genetic distances and high bootstrap support (or posterior probabilities). The computational burden involved in this phylogenetic threshold approach is a major drawback, especially when a large number of sequences are being considered. In addition, this method requires a skilled user to specify the appropriate threshold values which may vary widely depending on the application.ResultsThis paper presents the Gap Procedure, a distance-based clustering algorithm for the classification of DNA sequences sampled from individuals infected with the human immunodeficiency virus type 1 (HIV-1). Our heuristic algorithm bypasses the need for phylogenetic reconstruction, thereby supporting the quick analysis of large genetic data sets. Moreover, this fully automated procedure relies on data-driven gaps in sorted pairwise distances to infer clusters, thus no user-specified threshold values are required. The clustering results obtained by the Gap Procedure on both real and simulated data, closely agree with those found using the threshold approach, while only requiring a fraction of the time to complete the analysis.ConclusionsApart from the dramatic gains in computational time, the Gap Procedure is highly effective in finding distinct groups of genetically similar sequences and obviates the need for subjective user-specified values. The clusters of genetically similar sequences returned by this procedure can be used to detect patterns in HIV-1 transmission and thereby aid in the prevention, treatment and containment of the disease.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-015-0791-x) contains supplementary material, which is available to authorized users.

Distance Measure to Cluster Spatiotemporal Scenarios for Strategic Air Traffic Management

Convective weather, recognized as the leading contributor to delays in the National Airspace System, causes demand–capacity imbalance in the airspace. Strategic air traffic management aims to resolve the imbalance for offnominal conditions (for example, convective weather events) through redistributing flows and resources at the strategic timeframe (2–15 h in advance). As a component of scenario-driven strategic air traffic management, this paper develops a multiresolution spatiotemporal distance measure that, combined with standard distance-based clustering algorithms, can be used to group a wide range of possible weather-impact scenarios into a few representative clusters to facilitate corresponding management strategy design. Motivations of this new distance measure, its generation algorithm, and parameter impact analysis are described in detail to facilitate practical implementation and subsequent use in decision support. This multiresolution spatiotemporal distance measure not only addresses the needs of scenario-driven strategic air traffic management but also generally applies to broad data-driven decision support applications that involve large-scale physical processes of spatiotemporal spread dynamics, as the measure captures the similarity between scenarios of spatiotemporal spread patterns of varying shape, size, location, and intensity; corrects the “boundary effects”; and is flexible to a variety of data features in temporal and spatial dimensions. Performance evaluation is conducted through comparative studies and sensitivity analysis, using real weather-impact datasets.

Distance Based Clustering Algorithm for Energy Efficiency in Wireless Sensor Networks

In wireless sensor networks Energy-efficient routing is an important issue due to the limited battery power within the network, Energy consumption is one of the important performance factors. Specifically for the election of cluster head selection and distance between the cluster head node and base station. The main objective of this proposed system is to reduce the energy consumption and prolong the network lifetime. This paper introduces a new clustering algorithm for energy efficient routing based on a cluster head selection

Mining Data for Traffic Detection System Using GPS Enable Mobile Phone in Mobile Cloud Infrastructure

The increasing need for traffic detection system has become a vital area in both developing and developed countries. However, it is more important to get the accurate and valuable data to give the better result about traffic condition. For this reason, this paper proposes an approach of tracking traffic data as cheap as possible in terms of communication, computation and energy efficient ways by using mobile phone network. This system gives the information of which vehicles are running on which location and how much speed for the Traffic Detection System. The GPS sensor of mobile device will be mainly utilized to guess a user’s transportation mode, then it integrates cloud environment to enhance the limitation of mobile device, such as storage, energy and computing power. This system includes three main components: Client Interface, Server process and Cloud Storage. Some tasks are carried out on the Client. Therefore, it greatly reduces the bottleneck situation on Server side in efficient way. Most of tasks are executed on the Server and history data are stored on the Cloud Storage. Moreover, the paper mainly uses the distance based clustering algorithm in grouping mobile devices on the same bus to get the accurate data.

Sweep-Hyperplane Clustering Algorithm Using Dynamic Model

Clustering is one of the better known unsupervised learning methods with the aim of discovering structures in the data. This paper presents a distance-based Sweep-Hyperplane Clustering Algorithm (SHCA), which uses sweep-hyperplanes to quickly locate each point’s approximate nearest neighbourhood. Furthermore, a new distance-based dynamic model that is based on 2N-tree hierarchical space partitioning, extends SHCA’s capability for finding clusters that are not well-separated, with arbitrary shape and density. Experimental results on different synthetic and real multidimensional datasets that are large and noisy demonstrate the effectiveness of the proposed algorithm.

Sketched symbol recognition using Latent-Dynamic Conditional Random Fields and distance-based clustering

In this paper we propose a two-stage method for recognizing sketched symbols that combine the use of a discriminative model, for labeling symbol strokes and a distance-based clustering algorithm, for grouping the labels belonging to the same symbol. In the first stage, we employ Latent-Dynamic Conditional Random Field (LDCRF), a discriminative model able to analyze the features of unsegmented sequences of strokes by taking into account spatio-temporal information, and to classify the symbol parts by considering contextual information. In the second stage, the labels obtained from LDCRF are grouped into symbol labels by using a distance-based clustering algorithm which takes into account the geometric relationships among strokes. The effectiveness of our method has been evaluated on the domain of electric circuit diagrams achieving accuracy values varying between 81.3% and 91.0%.

A novel approach for distance-based semi-supervised clustering using functional link neural network

Semi-supervised clustering is gaining importance these days since neither supervised nor unsupervised learning methods in a stand-alone manner provide satisfactory results. Existing semi-supervised clustering techniques are mostly based on pair-wise constraints, which could be misleading. These semi-supervised clustering algorithms also fail to address the problem of dealing with attributes having different weights. In most of the real-life applications, all attributes do not have equal importance and hence same weights cannot be assigned for each attribute. In this paper, a novel distance-based semi-supervised clustering algorithm has been proposed, which uses functional link neural network (FLNN) for finding weights for attributes with small amount of labeled data for further use in parametric Minkowski's model for clustering. In FLNN, the nonlinearity is captured by enhancing the input using orthonormal basis functions. The effectiveness of the approach has been illustrated over a number of datasets taken from UCI machine learning repository. Comparative performance evaluation demonstrates that the proposed approach outperforms the existing semi-supervised clustering algorithms. The proposed approach has also been successfully used to cluster the crime locations and to find crime hot spots in India on the data provided by National Crime Records Bureau (NCRB).

Hamming Distance based Clustering Algorithm

Cluster analysis has been extensively used in machine learning and data mining to discover distribution patterns in the data. Clustering algorithms are generally based on a distance metric in order to partition the data into small groups such that data instances in the same group are more similar than the instances belonging to different groups. In this paper the authors have extended the concept of hamming distance for categorical data .As a data processing step they have transformed the data into binary representation. The authors have used proposed algorithm to group data points into clusters. The experiments are carried out on the data sets from UCI machine learning repository to analyze the performance study. They conclude by stating that this proposed algorithm shows promising result and can be extended to handle numeric as well as mixed data.

Including transcription factor information in the superparamagnetic clustering of microarray data

In this work, we modify the superparamagnetic clustering algorithm (SPC) by adding an extra weight to the interaction formula that considers which genes are regulated by the same transcription factor. With this modified algorithm which we call SPCTF, we analyze the Spellman et al. microarray data for cell cycle genes in yeast, and find clusters with a higher number of elements compared with those obtained with the SPC algorithm. Some of the incorporated genes by using SPCFT were not detected at first by Spellman et al. but were later identified by other studies, whereas several genes still remain unclassified. The clusters composed by unidentified genes were analyzed with MUSA, the motif finding using an unsupervised approach algorithm, and this allow us to select the clusters whose elements contain cell cycle transcription factor binding sites as clusters worthy of further experimental studies because they would probably lead to new cell cycle genes. Finally, our idea of introducing the available information about transcription factors to optimize the gene classification could be implemented for other distance-based clustering algorithms.

Average correlation clustering algorithm (ACCA) for grouping of co-regulated genes with similar pattern of variation in their expression values

Distance based clustering algorithms can group genes that show similar expression values under multiple experimental conditions. They are unable to identify a group of genes that have similar pattern of variation in their expression values. Previously we developed an algorithm called divisive correlation clustering algorithm (DCCA) to tackle this situation, which is based on the concept of correlation clustering. But this algorithm may also fail for certain cases. In order to overcome these situations, we propose a new clustering algorithm, called average correlation clustering algorithm (ACCA), which is able to produce better clustering solution than that produced by some others. ACCA is able to find groups of genes having more common transcription factors and similar pattern of variation in their expression values. Moreover, ACCA is more efficient than DCCA with respect to the time of execution. Like DCCA, we use the concept of correlation clustering concept introduced by Bansal et al. ACCA uses the correlation matrix in such a way that all genes in a cluster have the highest average correlation values with the genes in that cluster. We have applied ACCA and some well-known conventional methods including DCCA to two artificial and nine gene expression datasets, and compared the performance of the algorithms. The clustering results of ACCA are found to be more significantly relevant to the biological annotations than those of the other methods. Analysis of the results show the superiority of ACCA over some others in determining a group of genes having more common transcription factors and with similar pattern of variation in their expression profiles. Availability of the software: The software has been developed using C and Visual Basic languages, and can be executed on the Microsoft Windows platforms. The software may be downloaded as a zip file from http://www.isical.ac.in/~rajat. Then it needs to be installed. Two word files (included in the zip file) need to be consulted before installation and execution of the software.

Optimizing Cluster Heads for Energy Efficiency in Large-Scale Heterogeneous Wireless Sensor Networks

Many complex sensor network applications require deploying a large number of inexpensive and small sensors in a vast geographical region to achieve quality through quantity. Hierarchical clustering is generally considered as an efficient and scalable way to facilitate the management and operation of such large-scale networks and minimize the total energy consumption for prolonged lifetime. Judicious selection of cluster heads for data integration and communication is critical to the success of applications based on hierarchical sensor networks organized as layered clusters. We investigate the problem of selecting sensor nodes in a predeployed sensor network to be the cluster heads to minimize the total energy needed for data gathering. We rigorously derive an analytical formula to optimize the number of cluster heads in sensor networks under uniform node distribution, and propose a Distance-based Crowdedness Clustering algorithm to determine the cluster heads in sensor networks under general node distribution. The results from an extensive set of experiments on a large number of simulated sensor networks illustrate the performance superiority of the proposed solution over the clustering schemes based on k-means algorithm.

Shape-based clustering for 3D CAD objects: A comparative study of effectiveness

3D shape retrieval and clustering is of current interest in several different fields, including mechanical engineering. Several new shape representations for 3D objects are continuing to emerge. Most shape representations are embedded in a variety of feature spaces. However, some of the recently reported shape representations are embedded in arbitrary metric spaces, i.e. distance spaces, rather than in multi-dimensional feature space. For such representations, the only operations available on the data objects are distance calculations between the objects. In addition, some of the view-based representations are embedded in non-metric spaces where the representations and the corresponding distances do not follow the triangle inequality. For shape clustering applications, most existing algorithms assume the shape representations either to be embedded in a multi-dimensional feature space or a metric distance space, making it difficult to evaluate several shape representations that do not conform to these assumptions. Therefore, two different approaches were evaluated for using the distance features of a shape to obtain clustering results. In the first method, the original distances are transformed into feature space using a multi-dimensional scaling approach for use with K -means clustering. The second approach directly uses the original distances with a distance-based clustering algorithm. We compared the clustering effectiveness of these two approaches using a classified benchmark database of 3D models. The effect of using different shape descriptors and number of clusters was studied using four measures of clustering effectiveness. Several statistical methods, including the Rand Index and Mutual Information Index, were used to objectively evaluate the clustering efficacy.

Improving Personalization Solutions through Optimal Segmentation of Customer Bases

On the Web, where the search costs are low and the competition is just a mouse click away, it is crucial to segment the customers intelligently in order to offer more targeted and personalized products and services to them. Traditionally, customer segmentation is achieved using statistics-based methods that compute a set of statistics from the customer data and group customers into segments by applying distance-based clustering algorithms in the space of these statistics. In this paper, we present a direct grouping-based approach to computing customer segments that groups customers not based on computed statistics, but in terms of optimally combining transactional data of several customers to build a data mining model of customer behavior for each group. Then, building customer segments becomes a combinatorial optimization problem of finding the best partitioning of the customer base into disjoint groups. This paper shows that finding an optimal customer partition is NP-hard, proposes several suboptimal direct grouping segmentation methods, and empirically compares them among themselves, traditional statistics-based hierarchical and affinity propagation-based segmentation, and one-to-one methods across multiple experimental conditions. It is shown that the best direct grouping method significantly dominates the statistics-based and one-to-one approaches across most of the experimental conditions, while still being computationally tractable. It is also shown that the distribution of the sizes of customer segments generated by the best direct grouping method follows a power law distribution and that microsegmentation provides the best approach to personalization.

GAPS: A clustering method using a new point symmetry-based distance measure

In this paper, an evolutionary clustering technique is described that uses a new point symmetry-based distance measure. The algorithm is therefore able to detect both convex and non-convex clusters. Kd-tree based nearest neighbor search is used to reduce the complexity of finding the closest symmetric point. Adaptive mutation and crossover probabilities are used. The proposed GA with point symmetry (GAPS) distance based clustering algorithm is able to detect any type of clusters, irrespective of their geometrical shape and overlapping nature, as long as they possess the characteristic of symmetry. GAPS is compared with existing symmetry-based clustering technique SBKM, its modified version, and the well-known K -means algorithm. Sixteen data sets with widely varying characteristics are used to demonstrate its superiority. For real-life data sets, ANOVA and MANOVA statistical analyses are performed.