Clustering Criterion Research Articles

Unsupervised Spectral Clustering for Segmentation of Dynamic PET Images

Segmentation of dynamic PET images is needed to extract the time activity curves (TAC) of regions of interest (ROI). These TAC can be used in compartmental models for in vivo quantification of the radiotracer target. While unsupervised clustering methods have been proposed to segment PET sequences, they are often sensitive to initial conditions or favour convex shaped clusters. Kinetic spectral clustering (KSC) of dynamic PET images was recently proposed to handle arbitrary shaped clusters in the space in which they are identified. While improved results were obtained with KSC compared to three state of art methods, its use for clinical applications is still hindered by the manual setting of several parameters. In this paper, we develop an extension of KSC to automatically estimate the parameters involved in the method and to make it deterministic. First, a global search procedure is used to locate the optimal cluster centroids from the projected data. Then an unsupervised clustering criterion is tailored and used in a global optimization scheme to automatically estimate the scale parameter and the weighting factors involved in the proposed Automatic and Deterministic Kinetic Spectral Clustering (AD-KSC). We validate the method using GATE Monte Carlo simulations of dynamic numerical phantoms and present results on real dynamic images. The deterministic results obtained with AD-KSC agree well with those obtained with optimal manual parameterization of KSC, and improve the ROI identification compared to three other clustering methods. The proposed approach could have significant impact for quantification of dynamic PET images in molecular imaging studies.

Joint user clustering and resource allocation for device-to-device communication underlaying MU-MIMO cellular networks

In this paper, multiple device-to-device (D2D) communication underlaying cellular multiuser multiple inputs multiple outputs (MU-MIMO) systems is investigated. This type of communication can improve spectral efficiency to address future demand, but interference management, user clustering, and resource allocation are three key problems related to resource sharing. Interference alignment (IA) is proposed to better mitigate in-cluster interference compared with a multiplex scheme, and user clustering and resource allocation are jointly investigated using binary-integer programming. In addition to an exhaustive search for a maximum throughput, we propose a two-step suboptimal algorithm by reducing the search space and applying branch-and-bound searching (BBS). To further obtain a good trade-off between performance and complexity, we propose a novel algorithm based on distance-constrained criteria for user clustering. The simulation results show that the IA and multiplex schemes acquiring user clustering gains outperform the orthogonal scheme without user clustering. Besides, the proposed two-step and location-based algorithms achieve little losses compared with the optimal algorithm under low complexities.

A Novel Nested Circular Microphone Array and Subband Processing-Based System for Counting and DOA Estimation of Multiple Simultaneous Speakers

This paper addresses the topic of simultaneous speaker localization. The work is related to the generalized cross-correlation (GCC)-based methods for estimating the direction of multiple speakers. Considering the defects of GCC-based direction of arrival (DOA) estimation methods, we have applied several modifications to improve our previous subband processing-based system for the localization of simultaneous speakers. Three modifications have been presented in this paper. In the first step, the DOA estimation method is equipped with a front-end block that determines the number of speakers based on K-means clustering and silhouette criterion. This block provides the true number of speakers for the DOA estimator. Secondly, in order to eliminate the spatial aliasing, we propose a novel nested circular microphone array. In the proposed array design, each microphone pair is only used in appropriate subband according to its inter-microphone distance. In the third step, to overcome the weakness of GCC-phase transform (GCC-PHAT) in noisy and noisy-reverberant conditions, we propose a SNR estimation block. So, we can separate noisy and reverberant conditions and use PHAT filter for reverberant conditions and maximum likelihood filter for noisy situations. The proposed method has been evaluated on both simulated and real multi-speaker speech data in various environmental conditions and different number of speakers. Our evaluations in terms of DOA accuracy demonstrate the superiority of the proposed method compared to the fullband and baseline subband methods.

Spatial distribution patterns of plague hosts: point pattern analysis of the burrows of great gerbils in Kazakhstan.

AimThe spatial structure of a population can strongly influence the dynamics of infectious diseases, yet rarely is the underlying structure quantified. A case in point is plague, an infectious zoonotic disease caused by the bacterium Yersinia pestis. Plague dynamics within the Central Asian desert plague focus have been extensively modelled in recent years, but always with strong uniformity assumptions about the distribution of its primary reservoir host, the great gerbil (Rhombomys opimus). Yet, while clustering of this species’ burrows due to social or ecological processes could have potentially significant effects on model outcomes, there is currently nothing known about the spatial distribution of inhabited burrows. Here, we address this knowledge gap by describing key aspects of the spatial patterns of great gerbil burrows in Kazakhstan.LocationKazakhstan.MethodsBurrows were classified as either occupied or empty in 98 squares of four different sizes: 200 m (side length), 250 m, 500 m and 590–1020 m. We used Ripley's K statistic to determine whether and at what scale there was clustering of occupied burrows, and semi‐variograms to quantify spatial patterns in occupied burrows at scales of 250 m to 9 km.ResultsSignificant spatial clustering of occupied burrows occurred in 25% and 75% of squares of 500 m and 590–1020 m, respectively, but not in smaller squares. In clustered squares, the clustering criterion peaked around 250 m. Semi‐variograms showed that burrow density was auto‐correlated up to a distance of 7 km and occupied density up to 2.5 km.Main conclusionsThese results demonstrate that there is statistically significant spatial clustering of occupied burrows and that the uniformity assumptions of previous plague models should be reconsidered to assess its significance for plague transmission. This field evidence will allow for more realistic approaches to disease ecology models for both this system and for other structured host populations.

DSM-5 posttraumatic stress disorder symptoms associated with suicide behaviors in veterans.

A connection between suicidality and posttraumatic stress disorder (PTSD) has been consistently demonstrated; however, the underlying relationship between suicidality and PTSD remains unclear. The aim of this study was to examine patterns of DSM-5 PTSD symptom endorsement that differentiated veteran participants with and without a history of suicide behaviors. We enrolled 95 veterans, 32 of whom reported no suicide ideation (SI) or suicide attempts (SA). The 63 remaining participants reported a history of SI, with 28 of the 63 also reporting a historical SA. Participants completed a standardized diagnostic interview (Structured Clinical Interview for DSM-IV-TR; First, Spitzer, Gibbon, & Williams, 2002), structured interview of suicidal behaviors (Columbia-Suicide Severity Rating Scale; Posner et al., 2011), and selected clinical measures. Veterans who reported SI and/or SA were more likely to meet criteria for PTSD on DSM-5 than were veterans who reported neither SI nor SA. Participants who reported SA were more likely to meet criteria for clusters C and D. Finally, at the symptom level, those who reported SI were more likely to report experiencing feelings of alienation. Those who reported a SA were more likely to report avoidance of thoughts and feelings, inability to recall an important aspect of their trauma, persistent negative beliefs, diminished interest, and feelings of alienation. These findings suggest that targeting specific symptoms of PTSD may aid in treatment of suicidal thoughts and behaviors associated with PTSD.

An active contour model for the segmentation of images with intensity inhomogeneities and bias field estimation.

Intensity inhomogeneity causes many difficulties in image segmentation and the understanding of magnetic resonance (MR) images. Bias correction is an important method for addressing the intensity inhomogeneity of MR images before quantitative analysis. In this paper, a modified model is developed for segmenting images with intensity inhomogeneity and estimating the bias field simultaneously. In the modified model, a clustering criterion energy function is defined by considering the difference between the measured image and estimated image in local region. By using this difference in local region, the modified method can obtain accurate segmentation results and an accurate estimation of the bias field. The energy function is incorporated into a level set formulation with a level set regularization term, and the energy minimization is conducted by a level set evolution process. The proposed model first appeared as a two-phase model and then extended to a multi-phase one. The experimental results demonstrate the advantages of our model in terms of accuracy and insensitivity to the location of the initial contours. In particular, our method has been applied to various synthetic and real images with desirable results.

Th2‐high asthma: a heterogeneous asthma population?

high patients had cytokine mRNA levels above the 90 th percentile value in controls. Unsupervised hierarchical clustering was used to determine unique cytokine-based patient clusters. Results Cubic Clustering Criterion, pseudo F and t 2 statistics revealed a two- and a six-cluster model. The first cluster (n=23) was found in both models and consists of patients who present with an “IL-5-high and IL-17F-high” profile. Patients with an “IL-4- or IL-13-high” profile did not cluster in one single group. In the six-cluster model, the “IL-17F-low” group was divided into 5 separate clusters: “IL-5-high” profile (n=7), “IFN-g-high” profile (n=15), “IL-6- and/or TNF-high” profile (n=15), “IL-22-high” profile (n=15) and those patients that were low for all preceding cytokines (n=130; cluster 6). “IL-17F- and IL-5-high” patients had significantly lower FEV1 and higher sputum neutrophils. Patients that were only “IL-4or IL-13-high” (cluster 6) had highest FENO levels and sputum eosinophils. Conclusion Th2-high asthma can be subdivided in patients with “IL-5-high and IL-17F-high” asthma and those with “IL4- or IL-13-high” asthma. The inflammatory pattern is different between both groups. The former group is characterized by mixed granulocytic airway inflammation whereas the latter group consists of patients with eosinophilic airway inflammation.

Interactive and iterative visual clustering

This article proposes a semi-interactive system for visual data exploration using an iterative clustering that combines an automatic approach with an interactive one. We propose a framework to improve the interactivity between the user and the data analysis process, allowing him or her to participate actively in the iterative clustering tasks using a two-dimensional projection. Defining a cluster by its seed (center) and its limit, the proposed approach allows the user to modify the automated values or to define new seeds and the associated cluster limit himself or herself. The user can perform the clustering according to his or her visual perception manually and can also choose to let the automated approach find optimal seeds and then interact with the process to iterate the clustering process according to his or her visual perception and domain knowledge. Most of the evaluation criteria for clustering evaluate the complete clustering and not each cluster separately. In this article, we propose to adapt evaluation criteria to single clusters, allowing the users to evaluate their own clusters and perform the clustering iteratively until satisfaction. To evaluate our proposed approach, we conduct a user evaluation, where the users are asked to perform clustering interactively according to their visual perception and with the semi-interactive one. We also compare the obtained results with those of automated clustering. The quantitative results have shown that the cooperative approach can improve the clustering results in terms of accuracy.

Decision about criticality of power transformers using whitenization weight functions on DGA caution levels

Power transformers are the most significant as well as the major asset of any power system network. The condition monitoring and assessment is the main concern in transformer management activities. As a first information source, dissolved gases-in-oil analysis (DGA) is universally accepted. The assessment of dissolved gases is characteristically observed analogous to grey system analysis. Grey system theory is supportive to the cases, when less information about the system is available. The cluster of grey incidences and whitenization weight functions classifies the factors of same type, in order to simplify a complex system. Three caution levels of key gases specified in IEEE standards are utilized in this study, to whiten the weight functions. The whitenization weight function with lower measure is selected for caution level-1. However, whitenization weight functions with middle measure are preferred for level-2 and level-3. Several key gas samples of the equal rating transformers are collected from gas analyzer section and utilized in condition assessment computations. The test samples are verified with variable and equal weight clustering criteria. The criticality judgment of transformer with variable weight clustering successfully identifies the crucial elements amongst samples.

A fast clustering algorithm for data with a few labeled instances.

The diameter of a cluster is the maximum intracluster distance between pairs of instances within the same cluster, and the split of a cluster is the minimum distance between instances within the cluster and instances outside the cluster. Given a few labeled instances, this paper includes two aspects. First, we present a simple and fast clustering algorithm with the following property: if the ratio of the minimum split to the maximum diameter (RSD) of the optimal solution is greater than one, the algorithm returns optimal solutions for three clustering criteria. Second, we study the metric learning problem: learn a distance metric to make the RSD as large as possible. Compared with existing metric learning algorithms, one of our metric learning algorithms is computationally efficient: it is a linear programming model rather than a semidefinite programming model used by most of existing algorithms. We demonstrate empirically that the supervision and the learned metric can improve the clustering quality.

A model selection criterion for model-based clustering of annotated gene expression data.

In co-expression analyses of gene expression data, it is often of interest to interpret clusters of co-expressed genes with respect to a set of external information, such as a potentially incomplete list of functional properties for which a subset of genes may be annotated. Based on the framework of finite mixture models, we propose a model selection criterion that takes into account such external gene annotations, providing an efficient tool for selecting a relevant number of clusters and clustering model. This criterion, called the integrated completed annotated likelihood (ICAL), is defined by adding an entropy term to a penalized likelihood to measure the concordance between a clustering partition and the external annotation information. The ICAL leads to the choice of a model that is more easily interpretable with respect to the known functional gene annotations. We illustrate the interest of this model selection criterion in conjunction with Gaussian mixture models on simulated gene expression data and on real RNA-seq data.

Symmetry Based Automatic Evolution of Clusters: A New Approach to Data Clustering.

We present a multiobjective genetic clustering approach, in which data points are assigned to clusters based on new line symmetry distance. The proposed algorithm is called multiobjective line symmetry based genetic clustering (MOLGC). Two objective functions, first the Davies-Bouldin (DB) index and second the line symmetry distance based objective functions, are used. The proposed algorithm evolves near-optimal clustering solutions using multiple clustering criteria, without a priori knowledge of the actual number of clusters. The multiple randomized K dimensional (Kd) trees based nearest neighbor search is used to reduce the complexity of finding the closest symmetric points. Experimental results based on several artificial and real data sets show that proposed clustering algorithm can obtain optimal clustering solutions in terms of different cluster quality measures in comparison to existing SBKM and MOCK clustering algorithms.

Determining the management zones with hierarchic and non-hierarchic clustering methods

Delineation of the management zones of a field is commonly used in precision agriculture technology. There are many techniques used to identify management zones. The most used technique is k-means clustering, where the number of clusters is managed by the user. The paper deals with clustering the yield data and electromagnetic data of a 17 ha field using the Ward’s method followed by the k-means clustering method. The cubic clustering criterion was used to determine the number of clusters. Based on results, it can be concluded that it is beneficial to combine the k-means clustering method with the hierarchic method (Ward’s method).

Multi-scale permeability of deformable fibrous porous media

The contribution of fiber dynamics and clustering to the effective permeability in hierarchical fibrous media is poorly understood, due to the complex fluid–structure interactions taking place across fiber, yarn and textile scales. In this work, a two-dimensional model for fiber deformation subject to out-of-plane movement restrictions is derived for creeping flow conditions by analogy with non-Brownian suspensions of particles with confining potentials. This leads to a homogeneous Fokker–Planck equation in a phase space of fiber configurations, for the probability density function of the fiber displacements. A fiber clustering criterion is then defined using autoconvolution functions of the local probability densities, which yields the local change in fiber-scale permeability according to a topological description of the porous media instead of the typical geometric description. The resulting multi-scale hydrodynamic system is numerically solved by a coupled method, where the Stokes flow at yarn-scale is solved with a finite volume method and the mesoscopic model that recovers information from the fiber-scale is solved by a lattice Boltzmann method. The fiber-scale permeability is characterized in terms of porosity, dimensionless shear rate and dimensionless out-of-plane forces. When assessed in terms of a reduced viscosity related to Brinkman׳s closure for porous media, the mesoscopic model shows that deformable fibrous porous media qualitatively behave like dense particle suspensions. For low volume fractions a non-Newtonian reduced viscosity exhibiting shear-thinning and low- and high-shear plateaux is obtained. For high volume fractions and high shear rates the out-of-plane forces lead to shear thickening. The results on steady fiber-scale permeability are presented in the form of phase diagrams which show that in the typical range of parameters for textiles, the effective permeability of the deformable case can be up 60% lower than that of the rigid case due to the formation of fiber clusters.

Exploiting high level feature for dynamic textures recognition

In this paper, a novel framework is proposed for dynamic textures (DTs) recognition by learning a high level feature using deep neural network (DNN). The insight behind the method is that a DT appearing in different videos should share similar features, which can be learned for better recognition performance. Unlike many prior works only focus on low level or middle level features, we propose a novel high level feature learning method using DNN. Our goal is to construct a compact and discriminative semantic feature. The conventional bag of features approach using k-means is not semantically meaningful since the clustering criterion is based on appearance similarity. The proposed framework can effectively overcome the problem by capturing the semantic relations of the middle level by DNN. Extensive experiments with qualitative and quantitative results demonstrate the efficacy of our approach.

EdAl-2 (Educació en Alimentació) programme: reproducibility of a cluster randomised, interventional, primary-school-based study to induce healthier lifestyle activities in children

ObjectivesTo assess the reproducibility of an educational intervention EdAl-2 (Educació en Alimentació) programme in ‘Terres de l'Ebre’ (Spain), over 22 months, to improve lifestyles, including diet and physical activity (PA).DesignReproduction...

Unsupervised classification strategy utilizing an endmember extraction technique for airborne hyperspectral remotely sensed imagery

Remote sensing has become an important source of urban land-use/cover classification, and as a result of their high spatial and spectral resolution, airborne hyperspectral images have been widely used to distinguish different urban classes. However, the previous studies into the classification of urban environments have mainly focused on a supervised scenario, which is limited by the selection of training samples. An unsupervised classification strategy utilizing an endmember extraction technique for airborne hyperspectral imagery is proposed by a combination of endmember extraction and k-means classification. The number of endmembers of the hyperspectral image is first estimated with the hyperspectral signal subspace identification with the minimum error method, and then the simplex growing algorithm is used to extract the endmember spectra that represent the different latent materials in the hyperspectral imagery. These latent materials were further integrated into the predefined number of classes and the k-means classification method was utilized to obtain the final classification map. Different distance measures were experimentally used in the procedure of class integration and classification to investigate the impact of initial cluster centers and further clustering criterion. The proposed strategy was compared with three traditional unsupervised classification methods, k-means, fuzzy k-means, and ISODATA, with two airborne hyperspectral images. The experimental results demonstrate that the proposed approach is more robust and outperforms the other three classification methods, and, hence, provides an innovative perspective for implementing the unsupervised classification of airborne hyperspectral imagery.

Multilocal Search and Adaptive Niching Based Memetic Algorithm With a Consensus Criterion for Data Clustering

Clustering is deemed one of the most difficult and challenging problems in machine learning. In this paper, we propose a multilocal search and adaptive niching-based genetic algorithm with a consensus criterion for automatic data clustering. The proposed algorithm employs three local searches of different features in a sophisticated manner to efficiently exploit the decision space. Furthermore, we develop an adaptive niching method, which can dynamically adjust its parameter value depending on the problem instance as well as the search progress, and incorporate it into the proposed algorithm. The adaptation strategy is based on a newly devised population diversity index, which can be used to promote both genetic diversity and fitness. Consequently, diverged niches of high fitness can be formed and maintained in the population, making the approach well-suited to effective exploration of the complex decision space of clustering problems. The resulting algorithm has been used to optimize a consensus clustering criterion, which is suggested with the purpose of achieving reliable solutions. To evaluate the proposed algorithm, we have conducted a series of experiments on both synthetic and real data and compared it with other reported methods. The results show that our proposed algorithm can achieve superior performance, outperforming related methods.

Denis Dutton: Appreciation of the Man and Discussion of the Work

Despite being a trove of lively observations and stimulating ideas, The Art Instinct does not succeed in its stated purpose of showing that art has been evolutionarily adaptive. The book is more about aesthetic experience or response (pleasure and beauty) than art making or participation , and the author’s twelve “cluster criteria” are too general for understanding why a particular behavior (or behavioral predisposition) of art might have originated and evolved. Advocacy of the sexual-selection argument is inadequate: participation in the arts is good for everyone, not just a few (male) virtuosos.

Dynamic Point Clustering with Line Constraints for Moving Object Detection in DAS

In this letter, we propose a robust dynamic point clustering method for detecting moving objects in stereo image sequences, which is essential for collision detection in driver assistance system. If multiple objects with similar motions are located in close proximity, dynamic points from different moving objects may be clustered together when using the position and velocity as clustering criteria. To solve this problem, we apply a geometric constraint between dynamic points using line segments. Based on this constraint, we propose a variable K-nearest neighbor clustering method and three cost functions that are defined between line segments and points. The proposed method is verified experimentally in terms of its accuracy, and comparisons are also made with conventional methods that only utilize the positions and velocities of dynamic points.