Cluster Boundaries Research Articles

Clustering of velocities in a GPS network spanning the Sierra Nevada Block, the Northern Walker Lane Belt, and the Central Nevada Seismic Belt, California‐Nevada

AbstractThe deformation across the Sierra Nevada Block, the Walker Lane Belt, and the Central Nevada Seismic Belt (CNSB) between 38.5°N and 40.5°N has been analyzed by clustering GPS velocities to identify coherent blocks. Cluster analysis determines the number of clusters required and assigns the GPS stations to the proper clusters. The clusters are shown on a fault map by symbols located at the positions of the GPS stations, each symbol representing the cluster to which the velocity of that GPS station belongs. Fault systems that separate the clusters are readily identified on such a map. Four significant clusters are identified. Those clusters are strips separated by (from west to east) the Mohawk Valley‐Genoa fault system, the Pyramid Lake‐Wassuk fault system, and the Central Nevada Seismic Belt. The strain rates within the westernmost three clusters approximate simple right‐lateral shear (~13 nstrain/a) across vertical planes roughly parallel to the cluster boundaries. Clustering does not recognize the longitudinal segmentation of the Walker Lane Belt into domains dominated by either northwesterly trending, right‐lateral faults or northeasterly trending, left‐lateral faults.

Mapping world cultures: Cluster formation, sources and implications

This paper extends and builds on Ronen and Shenkar’s synthesized cultural clustering of countries based on similarity and dissimilarity in work-related attitudes. The new map uses an updated dataset, and expands coverage to world areas that were non-accessible at the time. Cluster boundaries are drawn empirically rather than intuitively, and the plot obtained is triple nested, indicating three levels of similarity across given country pairs. Also delineated are cluster adjacency and cluster cohesiveness, which vary from the highly cohesive Arab and Anglo clusters to the least cohesive Confucian and Far Eastern clusters. Exploring predictors of cluster formation, we draw on the ecocultural perspective and other inputs, and examine the combined role of language, religion, and geography in generating cluster formation. We find that these forces play a prominent yet complex role: for instance, the religion and language brought by the Spanish fail to create a singular, cohesive Latin American cluster akin to the Anglo cluster. The role of economic variables is similarly considered. Finally, comparing the current map with that of 1985, we find strong support for the divergence (vs convergence) argument. Implications for international business are delineated.

Deletion and Gene Expression Analyses Define the Paxilline Biosynthetic Gene Cluster in Penicillium paxilli

The indole-diterpene paxilline is an abundant secondary metabolite synthesized by Penicillium paxilli. In total, 21 genes have been identified at the PAX locus of which six have been previously confirmed to have a functional role in paxilline biosynthesis. A combination of bioinformatics, gene expression and targeted gene replacement analyses were used to define the boundaries of the PAX gene cluster. Targeted gene replacement identified seven genes, paxG, paxA, paxM, paxB, paxC, paxP and paxQ that were all required for paxilline production, with one additional gene, paxD, required for regular prenylation of the indole ring post paxilline synthesis. The two putative transcription factors, PP104 and PP105, were not co-regulated with the pax genes and based on targeted gene replacement, including the double knockout, did not have a role in paxilline production. The relationship of indole dimethylallyl transferases involved in prenylation of indole-diterpenes such as paxilline or lolitrem B, can be found as two disparate clades, not supported by prenylation type (e.g., regular or reverse). This paper provides insight into the P. paxilli indole-diterpene locus and reviews the recent advances identified in paxilline biosynthesis.

Large margin principle in hyperrectangle learning

In this paper, we propose a new meta learning approach to incorporate the large margin principle into hyperrectangle based learning. The goal of Large Margin Rectangle Learning (LMRL) is to combine the natural interpretability of hyperrectangle models like decision trees and rule learners with the risk minimization property related to the large margin principle. Our approach consists of two basic steps: supervised clustering and decision boundary creation. In the first step, we apply a supervised clustering algorithm to generate an initial rectangle based generalization of the training data. Subsequently, these labeled clusters are used to produce a large margin hyperrectangle model. Besides the overall approach, we also developed Large Margin Supervised Clustering (LMSC), an attempt to introduce the large margin principle directly into the supervised clustering process. Corresponding experiments not only provided empirical evidence for the supposed margin-accuracy relation, but also showed that LMRL performs equally well or better than compared decision tree and rule learner. Altogether, this new learning approach is a promising way to create more accurate interpretable models.

Propagation of dynamic nuclear polarization across the xenon cluster boundaries: Elucidation of the spin-diffusion bottleneck

Earlier Dynamic Nuclear Polarization (DNP) experiments with frozen xenon/1-propanol/trityl mixtures have demonstrated spontaneous formation of pure xenon clusters above 120K, enabling spectrally-resolved real-time measurements of 129Xe nuclear magnetization in the clusters and in the surrounding radical-rich matrix. A spin-diffusion bottleneck was postulated to explain the peculiar time evolution of 129Xe signals in the clusters as well as the apparent discontinuity of 129Xe polarization across the cluster boundaries. A self-contained ab initio model of nuclear spin diffusion in heterogeneous systems is developed here, incorporating the intrinsic T1 relaxation towards the temperature-dependent equilibrium polarization and the spin-diffusion coefficients based on the measured NMR line widths and the known atomic densities in each compartment. This simple model provides the physical basis for the observed spin-diffusion bottleneck and is in a good quantitative agreement with the earlier measurements. A simultaneous fit of the model to the time-dependent NMR data at two different DNP frequencies provides excellent estimates of the cluster size, the intrinsic sample temperature, and 129Xe T1 constants. The model was also applied to the NMR data acquired during relaxation towards the thermal equilibrium after the microwaves were turned off, to estimate T1 relaxation time constants inside and outside the clusters. Fitting the model to the data during and after DNP provides consistent estimates of the cluster size.

Non Metric Model Based on Rough Set Representation

The non metricmodel is a kind of clustering method in which belongingness or the membership grade of each object in each cluster is calculated directly from dissimilarities between objects and in which cluster centers are not used. The clustering field has recently begun to focus on rough set representation instead of fuzzy set representation. Conventional clustering algorithms classify a set of objects into clusters with clear boundaries, that is, one object must belong to one cluster. Many objects in the real world, however, belong to more than one cluster because cluster boundaries overlap each other. Fuzzy set representation of clusters makes it possible for each object to belong to more than one cluster. The fuzzy degree of membership may, however, be too descriptive for interpreting clustering results. Rough set representation handles such cases. Clustering based on rough sets could provide a solution that is less restrictive than conventional clustering and more descriptive than fuzzy clustering. This paper covers two types of Rough-set-based Non Metric model (RNM). One algorithm is the Roughset-based Hard Non Metric model (RHNM) and the other is the Rough-set-based Fuzzy Non Metric model (RFNM). In both algorithms, clusters are represented by rough sets and each cluster consists of lower and upper approximation. The effectiveness of proposed algorithms is evaluated through numerical examples.

Objective-Based Rough c-Means Clustering

Conventional clustering algorithms classify a set of objects into some clusters with clear boundaries, that is, one object must belong to one cluster. However, many objects belong to more than one cluster in real world since the boundaries of clusters generally overlap with each other. Fuzzy-set representation of clusters makes it possible for each object to belong to more than one cluster. On the other hand, the fuzzy degree is sometimes regarded as too descriptive for interpreting clustering results. Rough-set representation could deal with such cases. Clustering based on rough set could provide a solution that is less restrictive than conventional clustering and less descriptive than fuzzy clustering. This paper proposes a rough clustering algorithm which is based on optimization of an objective function and the calculation formula of cluster centers is the same as one by Lingras et al. Moreover, it shows effectiveness of our proposed clustering algorithm in comparison with other algorithms.

Clustering-boundary-detection algorithm based on center-of-gravity of neighborhood

The cluster boundary is a useful model, in order to identify the boundary effectively, according to the uneven distribution of data points int the epsilon neighborhood of boundary objects, this paper proposes a boundary detection algorithm S-BOUND. Firstly, all the points in the epsilon neighborhood of the data objects are projected onto the boundary of the convex hull of the neighborhood, and then calculate the center of gravity of the neighborhood. Finally, detect the boundary object according to the degree of deviation of the center of gravity of the neighborhood with the object. The experimental results show that the S-BOUND algorithm can accurately detect a variety of clustering boundary and remove the noises, the time of performance is also better. DOI: http://dx.doi.org/10.11591/telkomnika.v11i12.3620 Full Text: PDF

The genome of tolypocladium inflatum: evolution, organization, and expression of the cyclosporin biosynthetic gene cluster.

The ascomycete fungus Tolypocladium inflatum, a pathogen of beetle larvae, is best known as the producer of the immunosuppressant drug cyclosporin. The draft genome of T. inflatum strain NRRL 8044 (ATCC 34921), the isolate from which cyclosporin was first isolated, is presented along with comparative analyses of the biosynthesis of cyclosporin and other secondary metabolites in T. inflatum and related taxa. Phylogenomic analyses reveal previously undetected and complex patterns of homology between the nonribosomal peptide synthetase (NRPS) that encodes for cyclosporin synthetase (simA) and those of other secondary metabolites with activities against insects (e.g., beauvericin, destruxins, etc.), and demonstrate the roles of module duplication and gene fusion in diversification of NRPSs. The secondary metabolite gene cluster responsible for cyclosporin biosynthesis is described. In addition to genes necessary for cyclosporin biosynthesis, it harbors a gene for a cyclophilin, which is a member of a family of immunophilins known to bind cyclosporin. Comparative analyses support a lineage specific origin of the cyclosporin gene cluster rather than horizontal gene transfer from bacteria or other fungi. RNA-Seq transcriptome analyses in a cyclosporin-inducing medium delineate the boundaries of the cyclosporin cluster and reveal high levels of expression of the gene cluster cyclophilin. In medium containing insect hemolymph, weaker but significant upregulation of several genes within the cyclosporin cluster, including the highly expressed cyclophilin gene, was observed. T. inflatum also represents the first reference draft genome of Ophiocordycipitaceae, a third family of insect pathogenic fungi within the fungal order Hypocreales, and supports parallel and qualitatively distinct radiations of insect pathogens. The T. inflatum genome provides additional insight into the evolution and biosynthesis of cyclosporin and lays a foundation for further investigations of the role of secondary metabolite gene clusters and their metabolites in fungal biology.

The role of chromosomal rearrangements and geographical barriers in the divergence of lineages in a South American subterranean rodent (Rodentia: Ctenomyidae: Ctenomys minutus)

Identifying factors and the extent of their roles in the differentiation of populations is of great importance for understanding the evolutionary process in which a species is involved. Ctenomys minutus is a highly karyotype-polymorphic subterranean rodent, with diploid numbers ranging from 42 to 50 and autosomal arm numbers (ANs) ranging from 68 to 80, comprising a total of 45 karyotypes described so far. This species inhabits the southern Brazilian coastal plain, which has a complex geological history, with several potential geographical barriers acting on different time scales. We assessed the geographical genetic structure of C. minutus, examining 340 individuals over the entire distributional range and using information from chromosomal rearrangements, mitochondrial DNA (mtDNA) sequences and 14 microsatellite loci. The mtDNA results revealed seven main haplogroups, with the most recent common ancestors dating from the Pleistocene, whereas clustering methods defined 12 populations. Some boundaries of mtDNA haplogroups and population clusters can be associated with potential geographical barriers to gene flow. The isolation-by-distance pattern also has an important role in fine-scale genetic differentiation, which is strengthened by the narrowness of the coastal plain and by common features of subterranean rodents (that is, small fragmented populations and low dispersal rates), which limit gene flow among populations. A step-by-step mechanism of chromosomal evolution can be suggested for this species, mainly associated with the metapopulation structure, genetic drift and the geographical features of the southern Brazilian coastal plain. However, chromosomal variations have no or very little role in the diversification of C. minutus populations.

SVStream: A Support Vector-Based Algorithm for Clustering Data Streams

In this paper, we propose a novel data stream clustering algorithm, termed SVStream, which is based on support vector domain description and support vector clustering. In the proposed algorithm, the data elements of a stream are mapped into a kernel space, and the support vectors are used as the summary information of the historical elements to construct cluster boundaries of arbitrary shape. To adapt to both dramatic and gradual changes, multiple spheres are dynamically maintained, each describing the corresponding data domain presented in the data stream. By allowing for bounded support vectors (BSVs), the proposed SVStream algorithm is capable of identifying overlapping clusters. A BSV decaying mechanism is designed to automatically detect and remove outliers (noise). We perform experiments over synthetic and real data streams, with the overlapping, evolving, and noise situations taken into consideration. Comparison results with state-of-the-art data stream clustering methods demonstrate the effectiveness and efficiency of the proposed method.

Cluster boundary detection technology for categorical data

With the wide application of categorical-attribute dataset,the demand of obtaining the cluster boundary of categorical-attribute dataset becomes more and more urgent.In order to get cluster boundaries,a categorical-attribute data boundary detection algorithm: CBORDER(Categorical dataset BORDER detection algorithm) was proposed.In this algorithm,firstly,this paper initialized the center of cluster by using random allocation and utilizing boundary-degree to partition clusters;at the same time,the evidence of captured boundary records was got.Then,based on the evidence accumulation,the above procedure was executed repeatedly to acquire the boundaries of clusters at the end.The experimental results demonstrate that CBORDER can effectively detect the boundaries of the high-dimensional categorical data.

Distance-weighted city growth

Urban agglomerations exhibit complex emergent features of which Zipf's law, i.e., a power-law size distribution, and fractality may be regarded as the most prominent ones. We propose a simplistic model for the generation of citylike structures which is solely based on the assumption that growth is more likely to take place close to inhabited space. The model involves one parameter which is an exponent determining how strongly the attraction decays with the distance. In addition, the model is run iteratively so that existing clusters can grow (together) and new ones can emerge. The model is capable of reproducing the size distribution and the fractality of the boundary of the largest cluster. Although the power-law distribution depends on both, the imposed exponent and the iteration, the fractality seems to be independent of the former and only depends on the latter. Analyzing land-cover data, we estimate the parameter-value γ≈2.5 for Paris and its surroundings.

Unsupervised Spatiotemporal Analysis of FMRI Data Using Graph-Based Visualizations of Self-Organizing Maps

We present novel graph-based visualizations of self-organizing maps for unsupervised functional magnetic resonance imaging (fMRI) analysis. A self-organizing map is an artificial neural network model that transforms high-dimensional data into a low-dimensional (often a 2-D) map using unsupervised learning. However, a postprocessing scheme is necessary to correctly interpret similarity between neighboring node prototypes (feature vectors) on the output map and delineate clusters and features of interest in the data. In this paper, we used graph-based visualizations to capture fMRI data features based upon 1) the distribution of data across the receptive fields of the prototypes (density-based connectivity); and 2) temporal similarities (correlations) between the prototypes (correlation-based connectivity). We applied this approach to identify task-related brain areas in an fMRI reaction time experiment involving a visuo-manual response task, and we correlated the time-to-peak of the fMRI responses in these areas with reaction time. Visualization of self-organizing maps outperformed independent component analysis and voxelwise univariate linear regression analysis in identifying and classifying relevant brain regions. We conclude that the graph-based visualizations of self-organizing maps help in advanced visualization of cluster boundaries in fMRI data enabling the separation of regions with small differences in the timings of their brain responses.

Clustering of GPS velocities in the Mojave Block, southeastern California

AbstractWe find subdivisions within the Mojave Block using cluster analysis to identify groupings in the velocities observed at GPS stations there. The clusters are represented on a fault map by symbols located at the positions of the GPS stations, each symbol representing the cluster to which the velocity of that GPS station belongs. Fault systems that separate the clusters are readily identified on such a map. The most significant representation as judged by the gap test involves 4 clusters within the Mojave Block. The fault systems bounding the clusters from east to west are 1) the faults defining the eastern boundary of the Northeast Mojave Domain extended southward to connect to the Hector Mine rupture, 2) the Calico‐Paradise fault system, 3) the Landers‐Blackwater fault system, and 4) the Helendale‐Lockhart fault system. This division of the Mojave Block is very similar to that proposed by Meade and Hager []. However, no cluster boundary coincides with the Garlock Fault, the northern boundary of the Mojave Block. Rather, the clusters appear to continue without interruption from the Mojave Block north into the southern Walker Lane Belt, similar to the continuity across the Garlock Fault of the shear zone along the Blackwater‐Little Lake fault system observed by Peltzer et al. []. Mapped traces of individual faults in the Mojave Block terminate within the block and do not continue across the Garlock Fault [Dokka and Travis, ].

Asymptotics of the spectrum of the hydrogen atom in a magnetic field near the lower boundaries of spectral clusters

We investigate the second-order Zeeman effect in a magnetic field using irreducible representations of an algebra with the Karasev - Novikova quadratic commutation relations. To each such representation there corresponds a spectral cluster near the energy level of the unperturbed hydrogen atom. Using this model as an example, we describe a general method for constructing asymptotic solutions near the boundaries of spectral clusters based on a new integral representation. We also study the problem of computing quantum averages near the lower boundaries of clusters.

A Hybrid Cluster-Based Target Tracking Protocol for Wireless Sensor Networks

Target tracking is a typical and important application of wireless sensor networks (WSNs). In consideration of the network scalability and energy efficiency for target tracking in large-scale WSNs, it has been employed as an effective solution by organizing the WSNs into clusters. However, tracking a moving target in cluster-based WSNs suffers a boundary problem when the target moves across or along the boundaries of clusters, as the static cluster membership prevents sensors in different clusters from sharing information. In this paper, we propose a novel mobility management protocol, called hybrid cluster-based target tracking (HCTT), which integrates on-demand dynamic clustering into a cluster-based WSN for target tracking. By constructing on-demand dynamic clusters at boundary regions, nodes from different static clusters that detect the target can temporarily share information, and the tracking task can be handed over smoothly from one static cluster to another. As the target moves, static clusters and on-demand dynamic clusters alternately manage the target tracking task. Simulation results show that the proposed protocol performs better in tracking the moving target when compared with other typical target tracking protocols.

Asymptotics of the spectrum and quantum averages of a perturbed resonant oscillator near the boundaries of spectral clusters

In the eigenvalue problem for a perturbed two-dimensional oscillator, we suggest a method for constructing asymptotic solutions near the boundaries of spectral clusters by means of a new integral representation and study the issue of calculating the average values of differential operators on the solutions near the boundaries of the clusters.

Evolution of cooperation in lattice population with adaptive interaction intensity

We study the evolutionary Prisoner’s Dilemma game among individuals endowed with adaptively interaction intensity. Individuals adjust their interaction intensity according to the rules “payoff increase-high intensity, payoff decrease-low intensity”: if an individual’s payoff increases compared with that in the previous generation, he raises his interaction intensity; otherwise, he reduces the probability of interaction. We find that if individuals can adjust their interaction intensity with a proper scale, cooperation can be promoted. Interestingly, individuals with low interaction intensity usually hold the boundary of cooperator cluster. Such spatial distribution can alleviate the exploitation from defectors to cooperators since the interaction between cooperators and defectors is weakened. We hope our work can yield some insight into investigation of the evolution of cooperation in structured population.

Molecular Modeling of Cell Adhesion Peptides on Hydroxyapatite and TiO<sub>2</sub> Surfaces: Implication in Biomedical Implant Devices

Molecular modeling as a tool in studying peptide-substrate interactions provides insight on peptide adsorption conformation, adsorption energy, and stability of the peptide-inorganic interface. This work investigates the hydration and interaction of cell-adhesion peptides, specifically RGD and YIGSR, with the hydroxyapatite surface and TiO2 surface in cluster and periodic boundary condition approaches. The comparison of adsorption energies of RGD and YIGSR on both Hydroxyapatite (HA) and TiO2 surfaces reveals the similarities in adsorption energy and orientation pattern of peptides on both surfaces. The models demonstrate that initial peptide orientation affects adsorption energy for both. YIGSR is consistently more strongly adsorbed to HA-(001) surfaces and steps than RGD for both the surfaces. In addition, RGD maintained its “hairpin”-like structure during adsorption on a flat HA-(001) surface, and a slightly “relaxed hairpin” structure on TiO2 (110) surface. Adsorption energies of RGD on TiO2 (110) surface are significantly more favorable compared to HA-(001) surface, suggesting potential role of TiO2 as biomedical implants when tissue regeneration occurs via cell signaling.