Feature Extraction and Similarity of Movement Detection during Sleep, Based on Higher Order Spectra and Entropy of the Actigraphy Signal: Results of the Hispanic Community Health Study/Study of Latinos

The aim of this work was to develop a new unsupervised exploratory method of characterizing feature extraction and detecting similarity of movement during sleep through actigraphy signals. We here propose some algorithms, based on signal bispectrum and bispectral entropy, to determine the unique features of independent actigraphy signals. Experiments were carried out on 20 randomly chosen actigraphy samples of the Hispanic Community Health Study/Study of Latinos (HCHS/SOL) database, with no information other than their aperiodicity. The Pearson correlation coefficient matrix and the histogram correlation matrix were computed to study the similarity of movements during sleep. The results obtained allowed us to explore the connections between certain sleep actigraphy patterns and certain pathologies.

A New Method of Porous Space Reconstruction Using Multipoint Histogram Technology

Pore-scale modeling is becoming a hot topic in overall reservoir characterization process. It is an important approach for revealing the flow behaviors in porous media and exploring unknown flow patterns at pore scale. Over the past few decades, many reconstruction methods have been proposed, and among them the simulated annealing method (SAM) is extensively tested and easier to program. However, SAM is usually based on the two-point probability function or linear-path function, which fails to capture much more information on the multipoint connectivity of various shapes. For this reason, a new reconstruction method is proposed to reproduce the characteristics of a two-dimensional (2D) thin section based on the multipoint histogram. First, the two-point correlation coefficient matrix will be introduced to determine an optimal unit configuration of a multipoint histogram. Second, five different types of seven-point unit configurations will be used to test the unit configuration selection algorithm. Third, the multipoint histogram technology is used for generating the porous space reconstruction based on the prior unit configuration with a different calculation of the objective function. Finally, the spatial connectivity, patterns reproduction, the local percolation theory (LPT), and hydraulic connectivity are used to compare with those of the reference models. The results show that the multipoint histogram technology can produce better multipoint connectivity information than SAM. The reconstructed system matches the training image very well, which reveals that the reconstruction captures the geometry and topology information of the training image, for instance, the shape and distribution of pore space. The seven-point unit configuration is enough to get the spatial characters of the training image. The quality of pattern reproduction of the reconstruction is assessed by computing the multipoint histogram, and the similarity is around 97.3%. Based on the LPT analysis, the multipoint histogram can describe the anticipated patterns of geological heterogeneities and reproduce the connectivity of pore media with a high degree of accuracy. The two-point correlation coefficient matrix and a new construction theory are proposed. The new construction theory provides a stable theory and technology guidance for the study of pore space development and multiphase fluid flow rule in the digital rock.

Multiple h-index: a new scientometric indicator

Purpose – This paper aims to evaluate some of the known scientific indexes by using virtual data and proposes a new index, named multiple h-index (mh-index), for removing the limits of these variants. Design/methodology/approach – Citation report for 40 researchers in Babol, Iran, was extracted from the Web of Science and entered in a checklist together with their scientific lifetimes and published ages of their papers. Some statistical analyses, especially exploratory factor analysis (EFA) and structural correlations, were done in SPSS 19. Findings – EFA revealed three factors with eigenvalues greater than 1 and explained variance of over 96 per cent in the studied indexes, including the mh-index. Factors 1, 2 and 3 explained 44.38, 28.19 and 23.48 of the variance in the correlation coefficient matrix, respectively. The m-index (with coefficient of 90 per cent) in Factor 1, a-index (with coefficient of 91 per cent) in Factor 2 and h- and h2-indexes (with coefficients of 93 per cent) in Factor 3 had the highest factor loadings. Correlation coefficients and related comparative diagrams showed that the mh-index is more accurate than the other nine variants in differentiating the scientific impact of researchers with the same h-index. Originality/value – As the studied variants could not satisfy all limits of the h-index, scientific society needs an index which accurately evaluates individual researcher’s scientific output. As the mh-index has some advantages over the other studied variants, it can be an appropriate alternative for them.

Choose the Best Coach Based on the Theory of Grey System

In order to select “the best all time college coach” across both genders and all possible sports, we establish a multi-hierarchy evaluation model based on the theory of grey system. Without affecting the reliability of the results, we analyze the distribution of every index to determine the initial screening standard, narrow target range and simplify the problem. Then, we use the theory of grey system to evaluate coaches. We first normalize all indexes to make them comparable. Second, we determine a set of the best indexes as the reference vector and obtain the correlation coefficient matrix. On the issue of the weight of all indexes, we introduce the concept of dynamic weight matrix by considering the distribution and relative size of them. After that, we obtain the value of the final evaluation from the correlation coefficient matrix and dynamic weight matrix .As for the influence of time on the evaluation, we regroup the data of coaches according to the time section for five years, which results in some statistical variables, and then we draw the trend graph of them. To sum up, the number of excellent coaches of every sport is increasing over time, but the average value and standard deviation of winning percentage is decreasing. The model has the adaptability on different sports by comparison. And because the data of women coaches is insufficient, the influence of genders on the evaluation needs to be researched further.

Robust Principal Component Analysis Algorithm Based on Nonparametric Correlation Coefficient Matrix

Principal component analysis algorithm is widely applied in solving a large number of dimension reduction problems for high-dimensional data in fields such as image identification and comprehensive evaluation due to its clear conception and simple application. The PCA based on covariance matrix and correlation coefficient matrix is very sensitive to abnormal values so that the data structure is distorted. As a result, there exist great deviations in the analysis results and it is hard to explain. The traditional PCA is remade in this paper based on non-parametric statistical ideas, nonparametric correlation coefficient and gray relational coefficient, in order to define three kinds of robustness evaluation criteria and objectively evaluate the robustness of the algorithm through numerical experiments. The results of the numerical experiments show that the new algorithm is of good robustness to noise pollution and the algorithm is of good applicability.

Correlation Coefficient Matrix Based PD Feature Dimensional Reduction

The abstraction of diagnostic feature from field condition monitoring data is a significant research challenge. A new dimension reduction method based on correlation coefficient matrix is proposed aimed at the high-dimension characteristic parameters in the process of pattern recognition for partial discharge in power transformer. The CCM is constructed by parameters extracted from partial discharge signature in power transformer. The parameters that have similar classification characters are reduced directed by the correlation analysis result. The reduced PD features are inputted to the pattern classifiers of probabilistic neural networks (PNN). The results show that the parameter dimension is reduced and the classifier construction is simplified, and the recognition effect is better than that of the traditional back propagation neural network (BPNN) in the condition of small samples.

Seasonal Distribution of Phytoplankton in Riwada Reservoir, Visakhapatnam, Andhra Pradesh, India

The present study deals with seasonal variations, correlation coefficient and biodiversity indices of phytoplankton during April 2011 to March 2012 in the Riwada reservoir, Visakhapatnam, Andhra Pradesh, India. Sampling was performed at five stations during pre-monsoon, monsoon and post monsoon. There were a total of 57 genera belonging to four major groups i.e., Chlorophyceae (27 genera), Bacillariophyceae (14 genera), Cyanophyceae (13 genera) and Euglenophyceae (three genera). Maximum and minimum total phytoplankton population and percentages were recorded at station three in pre monsoon and at station two during monsoon. The maximum and minimum species richness (Menhinick index R2) were found to be 1.29 at station one and 1.10 at station three respectively. Maximum and minimum species diversity (H1) were found at station four (3.98) and station two (3.71). Maximum species evenness was recorded at stations one, being four and five; minimum species evenness was recorded at station two. Correlation coefficient matrix indicated significant positive relationship with water temperature, pH, transparency, biological oxygen demand and chlorides, negative relationship with electric conductivity, total solids, total dissolved solids, total hardness, dissolved oxygen, nitrates, sulphates and phosphates of water. The diversity indices showed that the reservoir have a well balanced phytoplankton community.

Gene Expression Network Reconstruction by LEP Method Using Microarray Data

Gene expression network reconstruction using microarray data is widely studied aiming to investigate the behavior of a gene cluster simultaneously. Under the Gaussian assumption, the conditional dependence between genes in the network is fully described by the partial correlation coefficient matrix. Due to the high dimensionality and sparsity, we utilize the LEP method to estimate it in this paper. Compared to the existing methods, the LEP reaches the highest PPV with the sensitivity controlled at the satisfactory level. A set of gene expression data from the HapMap project is analyzed for illustration.