Clustering Index Research Articles

Online Fault Diagnosis in Industrial Processes Using Multimodel Exponential Discriminant Analysis Algorithm

Fault diagnosis is used to identify the fault cause of the online abnormality, which is crucial for efficient and optimal operation of industrial processes. Due to the time-varying characteristics of fault process, the historical fault data may consist of multiple patterns and cannot be described accurately using a single model, which may result in poor performance of conventional multivariable statistical methods. In this brief, a multimodel exponential discriminant analysis (MEDA) algorithm is proposed for solving the aforementioned fault diagnosis problem. First, the samples of each fault class are clustered into different subclasses using the fast search and find of density peaks algorithm and the proposed cluster index. Subsequently, the between- and within-subclass exponential covariance matrices are calculated based on the subclasses and then the multimodel exponential discriminant model can be developed. Finally, recursively update the cluster center and the multimodel exponential discriminant model until the developed subclasses can be well separated and better discriminant performance can be obtained. Besides, a probabilistic MEDA algorithm and its corresponding online probabilistic diagnosis are also described to develop fuzzy model, so that the fault classes whose subclasses have fuzzy boundaries can also be effectively diagnosed. The Tennessee Eastman process is used to validate the diagnosis performance of the proposed MEDA algorithm, and the experimental results illustrate that the proposed algorithm can efficiently diagnose different classes of faults.

Performance analysis of clustering internal validation indexes with asymmetric clusters

The present work evaluates the performance of a set of internal clustering indexes in artificial and real data sets regarding a specific structural characteristic. In particular, it deals with data sets whose clusters present asymmetric characteristics in their geometries. With this objective, the concept of symmetry is formalized with respect to the axes of maximum variance of the clusters by the definition of a new index. Then a novel methodology is proposed to evaluate the performance of the internal indexes for crisp clustering in sets with this specific structural characteristic. The new defined index is combined with the correlation analysis, allowing to evaluate dynamically the performance of 11 internal indexes well known in the literature, and of a recently proposed index, the Representative Tree Index (RTI). In this way, this methodology allows us to measure not only the absolute error of the indices in relation to a particular configuration of clusters, but also the degree to which the structural characteristic of interest influences the performance of the index, obtaining a more generic understanding of their behaviors.

Variations and Phenetic Analysis of Peanut Cultivars (Arachis hypogaea L.) Based on Morphological Characteristics

Peanut (Arachis hypogaea L.) is a food commodity that is widely cultivated in Indonesia. At present there has been no analysis of the relationship between peanut cultivars with phenetic methods based on the morphological properties of the plants. Four cultivars of Arachis hypogaea L. Tuban, Talam 1, Talam 2, and Talam 3 used in this research. Morphological characters data was analyzed by description to construct identification key. Similarity index was counted by Simple Matching Coefficient (SSm) formula based on morphological scoring. Cluster analysis was conducted by UPGMA (Unweighted Pair Group Methods using Arithmetic Averages) method to construct dendrogram. PCA (Principal Component Analysis) were performed to defined role of each morphological character in grouping of accessions with MVSP (Multivariate Statistical Program) v. 3.1 software. The dendrogram showed that four cultivars of Arachis hypogaea L. divided into two main clusters, 4 sub-clusters. The similarity index of clusters is 0.85%.

Independent component analysis based algorithms for high-density electromyogram decomposition: Systematic evaluation through simulation

Motor unit activities provide important theoretical and clinical insights regarding different aspects of neuromuscular control. Based on high-density electromyogram (HD EMG) recordings, we systematically evaluated the performance of three independent component analysis (ICA)-based EMG decomposition algorithms (Infomax, FastICA and RobustICA). The algorithms were tested on simulated HD EMG signals with a range of muscle contraction levels and with a range of signal quality. Our results showed that all the three algorithms can output accurate (85%–100%) motor unit discharge timings. Specifically, the RobustICA consistently showed high decomposition accuracy among the three algorithms under a variety of signal conditions, especially with a low signal quality and varying contraction levels. But the yield of decomposition of RobustICA tended to be low at high contraction levels. In contrast, FastICA tended to show the lowest accuracy, but can detect the largest number of motor units, especially at high contraction levels. Our results also showed that the computation time was similar for FastICA and RobustICA, which was shorter than Infomax. Additionally, the accuracy of each algorithm correlated moderately with the clustering index—the silhouette distance measure, and correlated strongly with the rate of agreement of the algorithm pairs. Overall, our findings provide guidance on selecting particular decomposition algorithms based on specific applications with different requirement on the accuracy/yield of the decomposition.

EF-Index: Determining number of clusters (K) to estimate number of segments (S) in an image

Estimation of number of segments in an image attracts a formidable interest among the research community. The number of segments in an image is estimated by calculating the number of clusters present in the pixels of that image. The present work offers an unsupervised method, named “Electrostatic Force Index (EF-Index)”, to estimate the number of clusters inherent in an image, reporting of which is very rare in literature. The proposed approach is inspired by Coulomb's law of electrostatics. The EF-Index explores the mutual influence of an arbitrary pixel on another, by considering them similar to point charges. Our proposed cluster indexing method, viz. EF-Index is capable of determining the number of clusters present in an image. It has a strong resemblance to the way the electrostatic force is operative between a pair of static point charges in a closed system as per Coulomb's principle. In order to justify the effectiveness of the proposed approach, we have compared EF-Index of a given image with DB-Index, I-Index, CVNN-Index, DOE-AND-SCA and Sym-Index of the same image. Experimental results show that EF-Index is same as other state-of-the-art indices, whereas EF-Index does not require any clustering algorithm. To establish the applicability of the EF-Index, the same is applied for image segmentation considering Berkeley Segmentation Dataset and Stanford Background Dataset. We observe the results obtained conform to the ground truth and results achieved by applying existing well-established segmentation techniques on the same datasets. The efficacy of the proposed approach is further substantiated in terms of its reduced computational overhead in comparison to the state-of-the-art segmentation algorithms.

Dynamic Equivalent Modeling of Motors Based on Improved Hierarchical Clustering Algorithm

As the most important part of the power system load, the dynamic characteristics of motors are very important for power system simulation. The calculation is very large if all motors are modeled respectively. In order to reduce the amount of calculations, a dynamic equivalent modeling method of motors based on the improved hierarchical clustering algorithm is proposed in this paper. The two state parameters, the load rate and the product of inertia time constant and rotor resistance, are taken as clustering indexes. The improved hierarchical clustering algorithm is used to cluster dynamically the motors. And with the pseudo t2 statistics, it can obtain the best clustering results. Then it establishes the dynamic multi-machine equivalent model after each motor group equivalent. With the example system, the simulation results show that the proposed dynamic equivalent method can keep the dynamic characteristics of models and improve the equivalent accuracy.

Placing social types through prosodic variation: An investigation of spatial meanings in Mainland China

This study examines how prosodic features evoke the spacial aspects of interactional meanings of well-known social types in Mainland China. Prosodic features (duration, pitch, voice quality) of the scripted performances of 18 prominent social types in China were measured acoustically and grouped by cluster analysis. Commonalities among types within each group were identified through a detailed analysis of meta-linguistic commentary collected from the internet. This paper focuses on three meaningful clusters: powerful bureaucratic types, disembodied voices, and “in-your-face” types. Members of each cluster share prosodic combinations and social profiles. More importantly, character types within each cluster index a specific interactional locale. Appropriation of their associated features could reproduce the social dynamics that is typical in that locale. The results highlight the situated use of sociolinguistic variables, and show that the prosodic features pattern structurally in the performances while indexing the historical-spatial settings of social interactions. This paper also considers place as an interactional and relational product of meaning making by these prosodic features.

An improved form of the ant lion optimization algorithm for image clustering problems

Thispaperproposesanimprovedformoftheantlionoptimizationalgorithm(IALO)tosolveimageclustering problem. The improvement of the algorithm was made using a new boundary decreasing procedure. Moreover, a recently proposed objective function for image clustering in the literature was also improved to obtain well-separated clusters while minimizing the intracluster distances. In order to accurately demonstrate the performances of the proposed methods, firstly, twenty-three benchmark functions were solved with IALO and the results were compared with the ALO and a chaos-based ALO algorithm from the literature. Secondly, four benchmark images were clustered by IALO and the obtained results were compared with the results of particle swarm optimization, artificial bee colony, genetic, and K- means algorithms. Lastly, IALO, ALO, and the chaos-based ALO algorithm were compared in terms of image clustering by using the proposed objective function for three benchmark images. The comparison was made for the objective function values, the separateness and compactness properties of the clusters and also for two clustering indexes Davies– Bouldin and Xie–Beni. The results showed that the proposed boundary decreasing procedure increased the performance of the IALO algorithm, and also the IALO algorithm with the proposed objective function obtained very competitive results in terms of image clustering.

The spatial evolution of young massive clusters

Context. There are a number of methods that identify stellar sub-structure in star forming regions, but these do not quantify the degree of association of individual stars – something which is required if we are to better understand the mechanisms and physical processes that dictate structure.Aims. We present the new novel statistical clustering tool “INDICATE” which assesses and quantifies the degree of spatial clustering of each object in a dataset, discuss its applications as a tracer of morphological stellar features in star forming regions, and to look for these features in the Carina Nebula (NGC 3372).Methods. We employ a nearest neighbour approach to quantitatively compare the spatial distribution in the local neighbourhood of an object with that expected in an evenly spaced uniform (i.e. definitively non-clustered) field. Each object is assigned a clustering index (“I”) value, which is a quantitative measure of its clustering tendency. We have calibrated our tool against random distributions to aid interpretation and identification of significantIvalues.Results. Using INDICATE we successfully recover known stellar structure of the Carina Nebula, including the young Trumpler 14-16, Treasure Chest and Bochum 11 clusters. Four sub-clusters contain no, or very few, stars with a degree of association above random which suggests these sub-clusters may be fluctuations in the field rather than real clusters. In addition we find: (1) Stars in the NW and SE regions have significantly different clustering tendencies, which is reflective of differences in the apparent star formation activity in these regions. Further study is required to ascertain the physical origin of the difference; (2) The different clustering properties between the NW and SE regions are also seen for OB stars and are even more pronounced; (3) There are no signatures of classical mass segregation present in the SE region – massive stars here are not spatially concentrated together above random; (4) Stellar concentrations are more frequent around massive stars than typical for the general population, particularly in the Tr14 cluster; (5) There is a relation between the concentration of OB stars and the concentration of (lower mass) stars around OB stars in the centrally concentrated Tr14 and Tr15, but no such relation exists in Tr16. We conclude this is due to the highly sub-structured nature of Tr16.Conclusions. INDICATE is a powerful new tool employing a novel approach to quantify the clustering tendencies of individual objects in a dataset within a user-defined parameter space. As such it can be used in a wide array of data analysis applications. In this paper we have discussed and demonstrated its application to trace morphological features of young massive clusters.

Phenotypic, Hormonal, and Genomic Variation Among Vitis vinifera Clones With Different Cluster Compactness and Reproductive Performance.

Previous studies showed that the number of berries is a major component of the compactness level of the grapevine clusters. Variation in number of fruits is regulated by events occurring in the fruitset, but also before during the flower formation and pollination, through factors like the initial number of flowers or the gametic viability. Therefore, the identification of the genetic bases of this variation would provide an invaluable knowledge of the grapevine reproductive development and useful tools for managing yield and cluster compactness. We performed the phenotyping of four clones (two compact and two loose clones) of the Tempranillo cultivar with reproducible different levels of cluster compactness over seasons. Measures of reproductive performance included flower number per inflorescence, berry number per cluster, fruitset, coulure, and millerandage indices. Besides, their levels of several hormones during the inflorescence and flower development were determined, and their transcriptomes were evaluated at critical time points (just before the start and at the end of flowering). For some key reproductive traits, like number of berries per cluster and number of seeds per berry, clones bearing loose clusters showed differences with the compact clones and also differed from each other, indicating that each one follows different paths to produce loose clusters. Variation between clones was observed for abscisic acid and gibberellins levels at particular development stages, and differences in GAs could be related to phenotypic differences. Likewise, various changes between clones were found at the transcriptomic level, mostly just before the start of flowering. Several of the differentially expressed genes between one of the loose clones and the compact clones are known to be over-expressed in pollen, and many of them were related to cell wall modification processes or to the phenylpropanoids metabolism. We also found polymorphisms between clones in candidate genes that could be directly involved in the variation of the compactness level.

Patterns of tree diameter distributions in managed and unmanaged Abies alba Mill. and Fagus sylvatica L. forest patches

Temperate forests with shade-tolerant canopy tree species can develop vertical structures of varying complexity. Forests with Abies alba Mill. and Fagus sylvatica L. can be composed of one-, two-, and multi-storied patches and selection patches. A dominant view in forest ecology is that unmanaged forests tend to have greater structural heterogeneity than managed stands. Structural integrity, however, may differ among forest developmental stages. The main objective of this study was to compare the tree diameter complexity in managed and unmanaged patches during the early developmental stage.Data were collected between 2016 and 2018 in the Świętokrzyskie Mountains in Central Europe. The investigated tree communities were dominated by A. alba and F. sylvatica. Sample plots representing the growing-up developmental stage were randomly selected; of these, 30 plots were in managed stands, and 30 plots were in unmanaged forests. The diameter at breast height (DBH) distribution patterns were determined using hierarchical cluster analysis (HCA), clustering indices, and finite mixture models.Three main DBH distribution patterns were identified for the managed stands (K-A, K-B, and K-C). These patterns consisted of three or two sub-populations. The patterns represented structurally diversified patches composed of trees of all ages with multi-, three- or two-layered canopies and with intensive natural processes of regeneration. Two main DBH distribution patterns were identified for the unmanaged forests (S-A, and S-B). These patterns consisted of two clearly separated sub-populations. They are typical in patches with two-layered canopies, and the trees from the upper layer had a large share (40–60%). The distinguished DBH distribution patterns indicated there was greater tree size diversity in the managed stands than in the unmanaged forests. When comparing managed versus unmanaged patches, it is important to consider the developmental stage.

A social network analysis on immigrants and refugees access to services in the malaria elimination context

BackgroundThere has been significant progress in eliminating malaria in Iran. The aim of this study is to investigate the structure of inter-organizational collaboration networks in the field of unauthorized immigrants and refugees access to services in order to eliminate malaria.MethodsThis study employed social network analysis, in which nodes represented stakeholders associated with providing access of immigrants and refugees to services in the field of malaria elimination, and ties indicated the level of collaboration. This study adopted socio-centric analysis and the whole network was studied. In this regard, 12 districts of the malaria-endemic area in Iran were selected. Participants included 360 individuals (30 representatives of the organization/group in each district). The data were gathered by interview, using the levels of collaboration scale. UCINET 6 was used for data analysis. The indices of density, centralization, reciprocity, and clustering were investigated for each twelve network and at each level of collaboration.ResultsThe average density of the networks was 0.22 (SD: 0.04). In districts with a high incidence of imported malaria, the values of network density and centralization were high and the networks comprised of a larger connected component (less isolated clusters). There were significant correlations between density of network (r = 0.66, P = 0.02), degree centralization (r = 0.65, P = 0.02), betweenness centralization (r = 0.76, P = 0.004), and imported malaria cases. In general, the degree centrality and betweenness centrality of the organizations of health, district governor, and foreign immigrants’ affairs were higher. In all networks, 60% of the relationships were bilateral. At a higher level of collaboration, the centralization declined and reciprocity increased. The average of betweenness centralization index was 22.76 (SD = 3.88).ConclusionsHigher values of network indices in border districts and districts with more cases of imported malaria, in terms of density and centralization measures, can propose the hypothesis that higher preparedness against the issue and centralization of power can enable a better top-down outbreak management, which needs further investigations. Higher centrality of governmental organizations indicates the need for involving private, non-governmental organizations and representatives of immigrant and refugee groups. Recognition of the existing network structure can help the authorities increase access to malaria prevention, diagnosis, and treatment services among immigrants and refugees.

Indexes in Microsoft SQL Server

Indexes are the best apposite choice for quickly retrieving the records. This is nothing but cutting down the number of Disk IO. Instead of scanning the complete table for the results, we can decrease the number of IO's or page fetches using index structures such as B-Trees or Hash Indexes to retrieve the data faster. The most convenient way to consider an index is to think like a dictionary. It has words and its corresponding definitions against those words. The dictionary will have an index on because when we open a dictionary and we want to fetch its corresponding word quickly, then find its definition. The dictionary generally contains just a single index - an index ordered by word. When we modify any record and change the corresponding value of an indexed column in a clustered index, the database might require moving the entire row into a separately new position to maintain the rows in the sorted order. This action is essentially turned into an update query into a DELETE followed by an INSERT, and it decreases the performance of the query. The clustered index in the table can often be available on the primary key or a foreign key column because key values usually do not modify once a record is injected into the database.

Equivalent Modeling of Photovoltaic Power Station Based on Canopy-FCM Clustering Algorithm

Grid-connected photovoltaic power generation has become an effective way to utilize solar energy. In order to accurately analyze the dynamic characteristics of a grid-connected photovoltaic power station, an equivalent modeling method based on the Canopy-FCM clustering algorithm is proposed. Considering the lack of theoretical basis for the selection of the traditional clustering index, this paper deduces the transfer function of the grid-connected inverter control system based on the detailed model and proposes a new clustering index, which takes the zero-pole expression of the closed-loop transfer function as the prototype. In order to improve the fuzzy C-means clustering method, which is sensitive to the initial clustering centers and the outliers, a Canopy algorithm with lower computation cost is introduced. Before fuzzy C-means clustering analysis, the Canopy algorithm is used to determine the initial clustering centers and the number of clusters to improve the accuracy and efficiency of the clustering algorithm. The effectiveness of the proposed method is verified by simulation examples.

Equivalent Modeling of Photovoltaic Power Plant Based on Factor Analysis and Correlation Clustering

The equivalent model of photovoltaic (PV) power plant can greatly improve the model simulation speed on the basis of satisfying the simulation accuracy. However, in equivalent modeling of PV power plant, there are strong correlations and subjectivity between the selected initial variables, which will cause inaccurate equivalent results. Therefore, this paper proposes an equivalent modeling method of PV power plant based on factor analysis and correlation clustering. First, the factor analysis is used to reduce the dimensions of initial variables to eliminate the correlations between clustering indexes. The factor rotation and factor scores are adapted to mine the actual physical meaning of data, extracting the operation states of PV units, and eliminating the subjectivity of clustering indexes. Second, the correlation analysis and significance test are applied to analyze the extracted operation state indexes to obtain the correlations between PV units for clustering. Third, by the contour visualization, the correlations between PV units are qualitatively analyzed and the rationality of clustering results is judged. Finally, the equivalent parameters are calculated based on clustering results to establish the equivalent model of the PV power plant. The simulation results show that the proposed equivalent modeling method not only improves the simulation speed but also fit the simulation accuracy as well. The adaptability of the equivalent modeling method under different working conditions is also verified. The established equivalent model can be used to analyze the interaction between the PV power plant and power grid as well as the power system analysis with large-scale new energy, which is practical and popularized.

A Dynamic Equivalent Model for DFIG-Based Wind Farms

With the increasing scale of wind farms, the fault characteristics tend to be complex, which poses a technical challenge to establish the dynamic equivalent model of wind farms. In this paper, a dynamic equivalent method of DFIG-based wind farm based on the density peak clustering algorithm (DPCA) is presented. First, under an analysis of short-circuit current (SCC) in single doubly fed induction generator (DFIG), the clustering indexes are selected. Second, with the selected clustering indexes and DPCA, a more refined two-stage clustering of DFIGs in wind farm is carried out. Third, the units in the same cluster are equivalent to one unit, and then the dynamic equivalent model of the DFIG-based wind farm is established. Finally, the proposed method is validated through the MATLAB/SIMULINK-based simulation results, and the comparison results also show that the dynamic equivalent model proposed in this paper has a better performance than two other equivalent models. Moreover, another comparison between DPCA and K-means clustering algorithm is analyzed, and the result shows that DPCA has a better performance which provides a better choice for dynamic equivalence of wind farms.

Using Combined Descriptive and Predictive Methods of Data Mining for Coronary Artery Disease Prediction: a Case Study Approach

Heart disease is one of the major causes of morbidity in the world. Currently, large proportions of healthcare data are not processed properly, thus, failing to be effectively used for decision making purposes. The risk of heart disease may be predicted via investigation of heart disease risk factors coupled with data mining knowledge. This paper presents a model developed using combined descriptive and predictive techniques of data mining that aims to aid specialists in the healthcare system to effectively predict patients with Coronary Artery Disease (CAD). To achieve this objective, some clustering and classification techniques are used. First, the number of clusters are determined using clustering indexes. Next, some types of decision tree methods and Artificial Neural Network (ANN) are applied to each cluster in order to predict CAD patients. Finally, results obtained show that the C&RT decision tree method performs best on all data used in this study with 0.074 error. All data used in this study are real and are collected from a heart clinic database.

Phytosociological Survey of Sugarcane Crop Weeds in Different Agroecological Areas in Tucumán Province, Argentina

ABSTRACT: Weeds constitute the main constraint on the productivity of sugarcane agroecosystems in Argentina. This work aimed to analyze the floristic and phytosociological composition of sugarcane weed communities in different agroecological areas in Tucumán. Fifteen plots from each area were sampled 60 to 90 days after crop shooting, by throwing a 0.5 m x 0.5 m quadrat metallic frame six times in random directions in each of the plots. Species found within the frame were identified. Later, fresh and dry biomass weight were calculated per quadrat sample and species. Frequency, density, dominance, importance value index, Shannon-Wiener and Simpson species diversity index and Jaccard similarity index were calculated. The unweighted pair group method with arithmetic mean (UPGMA) was used to interpret similarity. Overall, 35 species were identified, 24 of which were dicotyledons and 11 monocotyledons. Eighteen of them were annual species, versus 17 which were perennial. Considering IVI, the most outstanding families were Poaceae, Cyperaceae, Euphorbiaceae, Amaranthaceae and Asteraceae. The results obtained in all the areas revealed the importance of the species Panicum maximum, Sorghum halepense, Cyperus rotundus, Cynodon dactylon and Euphorbia hyssopifolia. The areas with greater and less diversity were the Depressed Plain-Non-Saline Depressed Plain Subregion and the Subhumid-Humid Chaco-Pampean Plain, respectively. There was a 53% similarity between the Depressed Plain-Saline Depressed Plain and the Depressed Plain-Non-Saline Depressed Plain, and a 47% similarity between the Humid and Perhumid Foothills and the Subhumid-Humid Chaco-Pampean Plain. Hierarchical clustering and Jaccard index led to similar results.

Scalable keyword-based search and data manipulation on encrypted data

Cloud-based services increase productivity and reduce infrastructure cost industries which attract researchers and individuals outsource data to the remote cloud servers. However, cloud servers pose several security issues and the most prominent issue is privacy. Although encrypting the data before outsourcing preserves the data privacy, it does not support data usability such as searching with keywords on encrypted data. Searchable symmetric encryption (SSE) is found to be an efficient solution and in this paper, we construct an SSE method which enables search operation on encrypted data and allows various data operations on encrypted data. We perform document clustering and construct index tree-based secure search scheme. The index and queries are constructed with vector space model and are encrypted using secure k-NN computation method. We defined two encrypted searchable schemes based on two security models. We perform thorough experiments to justify the efficiency of the proposed model.

Correlation and path coefficient analysis in blackgram [Vigna mungo (L.) Hepper

Fourty five blackgram genotypes were studied including two check varieties (Pratap Urd-1 and PU-31) in randomized block design with three replications to study correlation and path coefficient analysis for 14 traits. Seed yield per plant was positively and significantly correlated with number of branches per plant, number of clusters per plant, number of pods per plant, number of pods per cluster, biological yield and harvest index at both phenotypic as well as genotypic level. The positive direct effect on seed yield per plant was exhibited by number of pods per plant, biological yield, harvest index, days to 50 per cent flowering, plant height and number of seeds per pod. Thus, seed yield per plant can be improved by practicing selection for number of pods per plant, biological yield, harvest index, days to 50 per cent flowering, plant height and number of seeds per pod.