Gaussian Process Regression Networks

We introduce a new regression framework, Gaussian process regression networks (GPRN), which combines the structural properties of Bayesian neural networks with the non-parametric flexibility of Gaussian processes. This model accommodates input dependent signal and noise correlations between multiple response variables, input dependent length-scales and amplitudes, and heavy-tailed predictive distributions. We derive both efficient Markov chain Monte Carlo and variational Bayes inference procedures for this model. We apply GPRN as a multiple output regression and multivariate volatility model, demonstrating substantially improved performance over eight popular multiple output (multi-task) Gaussian process models and three multivariate volatility models on benchmark datasets, including a 1000 dimensional gene expression dataset.

Relationship between El Nino and summer monsoon rainfall over Pakistan

The present paper deals with the influence of El Nino event on the summer monsoon rainfall over Pakistan. The correlation between monthly rainfall of summer monsoon season and bi-Monthly Multivariate ENSO Index (MEI) has been calculated to see the influence of El Nino on the summer monsoon rainfall. MEI is bimonthly ENSO Index pertaining to the period from first week of previous month to first week of the month under consideration. While study the correlation's with the ENSO events out side the Pacific Ocean MEI is more appropriate than other indices like Southern Oscillation Index (SOI) as MEI integrates complete information on ENSO viz. six oceanic and meteorological variables over the tropical Pacific. The results of the study show that there is a tendency of reduction in summer monsoon rainfall over Pakistan during El Nino years. The deficiency in % rainfall is statistically significant up to 90% level during July and September months. It is interesting to note that Pakistan receives more than normal rainfall during summer monsoon season in the immediate following year after the El Nino event. The correlation analysis is also performed on the summer monsoon months for individual provinces of Pakistan. All provinces receive deficient rainfall during monsoon months. The deficiency in rainfall over Punjab during all monsoon months is significant, whereas the deficiency in rainfall is significant during July and August over NWFP and Sindh respectively. No significant impact of El Nino on the summer monsoon rainfall over Baluchistan is observed.

Survey of plankton blooms and red tides and their interaction with environmental factors through monitoring satellite images

One of the most important marine ecological phenomena is red tide which is created by increasing of phytoplankton population, influenced by different factors such as climate condition changes, utrification hydrological factors and can leave sever and undesired ecological and economical effects behind itself in the case of durability. Coast line of Hormozgan is about 900km from east to west, within the range of geographical coordinates of 56 16 23.8, 26 58 8.8 to 54 34 5.33 and 26 34 32 eastern longitude and northern latitude, seven sampling stations were considered and sampled for a period of one year from October 2008 to October 2009. after the analysis of Satellite images, monthly, during the best time. In several stages, samplings were performed. In each station, three samples were collected for identification and determination of Bloom- creating species abundance. Cochlodinium polykrikoides was the species responsible for the discoloration which occurred at October 2008 in Hormozgan marine water. Environmental parameters such as sea surface temperature, pH, salinity, Dissolved Oxygen concentration, Total Dissolved Solids (T.D.S.), conductivity, nitrate, nitrite and phosphate and also chlorophyll a were measured and calculated. Kruscal Wallis test was used to compare the densities between different months, seasons and the studied stations. Mann-whitney test from Nonparametric Tests was used for couple comparison. Pearson correlation coefficient was used to determine the relationship between physical and chemical data set and the abundance of Cochlodinium polykrikoides. Multivariate Regression and analysis of variance (ANOVA) also were used to obtain the models and equations of red tide occurrence relationship, environmental parameters and nutrient data. The highest density was 26 million cells per liter in Qeshm station. A meaningful difference was observed between sampling months and seasons but there was no between sampling stations which indicates that in favorable conditions, the occurrence of this phenomenon by the studied species is probable. Regarding to β coefficients of nitrate, temperature, phosphate, Total Dissolvable Solutions (T.D.S) and pH these parameters are effective on the abundance of this species and red tide occurrence. Increase in these factors can represent the effects and outcomes of human activities and increase in marine pollution.

Differentiation of subspecies of asiatic striped squirrels (Tamiops swinhoei) (Milne-Edwards) (Rodentia : Sciuridae) in China with description of a new subspecies

Based on the pelage characteristics and results of multivariate and univariate analyses, a new subspecies is described in this study, and a taxonomic revision of Tamiops swinhoei from China is presented. In total, 123 specimens of Tamiops swinhoei were in

Geographic variation of Chevrier's field mouse (Apodemus chevrieri) (Milne-Edwards, 1868) (Muridae : Murinae) from southwestern China based on cranial morphometric variables

Geographic variation of Chevrier's field mouse (Apodemus chevrieri) (Milne-Edwards, 1868) (Muridae: Murinae) from southwestern China based on cranial morphometric variables. Zoological Studies 47(4): 393-401. A sample of 134 specimens of Apodemus chevrieri was investigated in the present study. Individuals were divided into male and female groups, and these were respectively subjected to multivariate analysis. Results indicated that 3 geographic populations of A. chevrieri inhabit southwestern China: a Sichuan population in western Sichuan Province; a northwestern Yunnan population ranging from northwestern Yunnan Province eastward to southern Sichuan Province; and a central Yunnan population in central Yunnan Province. In addition, a coefficient of difference analysis was performed among these 3 geographic populations. The results suggested that these 3 geographical populations of A. chevrieri belonged to 2 subspecies. Furthermore, we discuss the relationships of the subspecific differentiation of A. chevneri with changes in latitude in southwestern China.

滇池金线鲃形态度量学分析(Cypriniformes:Cyprinidae)

运用多变量形态度量学和框架结构形态学的方法,测量了采自滇池流域6个不同地点的149尾滇池金线鲃(Sinocyclocheilus grahami)的标本.应用多变量形态度量学分析,选择了20个框架结构性状和19个常规性状,进行主成成分分析.主成成分分析结果的散布图显示,6个不同地点的标本聚在一起没有分开,表明滇池金线鲃的各居群之间在可数性状和可量性状上均未表现出明显的差异,说明滇池金线鲃在形态上还未发生明显的种下分化.

Geographic variation of the Perny's Long-nosed squirrels (Dremomys pernyi) (Milne-Edwards, 1867) (Rodentia : Sciuridae) from southwestern China based on cranial morphometric variables

A sample of 114 specimens of Dremomys pernyi was investigated, 73 of which had intact skulls and were subjected to multivariate, coefficient of difference (C. D.), and cluster analyses. Results indicate that 4 subspecies (groups) of Dremomys pernyi inhabi

Bayesian classification of acoustical waveforms under environmental variability

We present in this paper a new multivariate probabilistic approach to Acoustic Pulse Recognition (APR) for tangible interface applications. This model uses Principle Component Analysis (PCA) in a probabilistic framework to classify tapping pulses with a high degree of variability. It was found that this model, achieves a higher robustness to pulse variability than simpler template matching methods, specifically when allowed to train on data containing high variability. © 2011 IEEE.

A COMPARATIVE-STUDY OF THE ELBOW JOINTS IN 5 SPECIES OF CHINESE MACAQUES

In this paper, we report the findings of a comparative study of the elbow joints of five species of macaque that inhabit China: Macaca assamensis, M. arctoides, M. mulatta, M. thibetana and M. nemestrina. Results of multivariate analyses of size-related variables and indices of the elbow joint suggested that the breadths of the ventral aspect of the trochlea and of the medial epicondyle of the humerus as well as indices describing the head of the radius are important factors for discriminating these species. The elbow joint of M. arctoides was most similar to that of M. thibetana, no doubt reflecting recency of common ancestry and similarity in terrestrial locomotion. The structures of the elbow joints in M. nemestrina and assamensis seemed more adapted to arboreal quadrupedalism. The elbow joint of M. mulatta, however, appears intermediate between the most terrestrial and the most arboreal forms.

COMPARISON OF MASTICATORY MORPHOLOGY BETWEEN RHINOPITHECUS-BIETI AND R-ROXELLANA

The masticatory apparatus for two endemic species of golden monkey in China, Rhinopithecus bieti and Rhinopithecus roxellana, were compared with those of macaques, Macaca and leaf monkeys, Presbytis. Multivariate analyses demonstrated that the two golden monkey species are distinct. Interspecies allometric analyses revealed that golden monkeys differ in their masticatory apparatus from both macaques and leaf monkeys. The prominent symphysial fusion, corpus, and sagittal condylar dimension of R. roxellana may produce efficient biting force on the incisal and posterior canine teeth, with the heavy reaction force barn on the temporomandibular joint. However, the well-developed bizygamatic width and mandibular height in R. bieti suggest that posterior canine function is similarly prominent in R. roxellana, while incisal function is not. (C) 1995 Wiley-Liss, Inc.

Metrical dental analysis on golden monkey (Rhinopithecus roxellana)

Dental variation in the Chinese golden monkey (Rhinopithecus roxellana) is here evaluated by univariate, bivariate, and multivariate analyses. Allometric analyses indicate that canines and P3s are positively, but other dimensions negatively scaled to mandible and maxilla, and to body size. With the exception of the mesiodistal dimensions of I-1 and M-3, and the buccolingual dimension of Pq, mandibular dental variables show similar scaling relative to body size. Analysis of residuals shows that males have significantly larger canine, P-3 and buccolingual dimensions of the postcanine teeth (M-2 and M-3) than females. A significant difference in shape between the sexes is found in the buccolingual dimension of the upper teeth, but not in the mandible. Unlike the situation in some other species, Female golden monkeys do nor exhibit relatively larger postcanine teeth than males, in fact, the reverse is true, especially for M(2)s and M(3)s. The fact that most of the dental variables show low negative allometry to body size might be related a cold environment that has led to the development of larger body size with I-educed energy loss. When the raw data are examined by Discriminant Function Analysis the sexes are clearly distinguishable.

Mandibular variation among Chinese macaques

In order to clarify the degree to which mandibular variation among Chinese macaques results from functional adaptation and phylogenetic inertia, 13 mandibular variables were analyzed by bivariate and multivariate techniques. The results indicate, not surprisingly, that the main differences in the mandible are associated with size. The study further implies that the variation between species is not closely associated with differences in functional adaptation even though the dietary and related differences are large compared to the situation in other macaques. The great variety in diet and related factors among Chinese macaques may not have yet resulted in a significant variation in the mandible. This may be because their radiation in Asia, though involving considerably greater differences in habitat, climate, and so on, has occurred more recently than for other macaque species in Southeast Asia. Mandibular variation between these species, therefore, is likely to be more closely tied to their immediate prior phylogenetic history. For example, the two stump-tailed macaques are closely similar and are also closely similar to the Assam species. Function in the mandible in these species is quite different. The results, therefore, seem to support the hypothesis that these three macaque species should be placed in a single species-group (sinica) as proposed by Delson [1980], Pan [1998], and Pan et al. [1998]. (C) 2002 Wiley-Liss, Inc.

Statistics anxiety, state anxiety during an examination, and academic achievement.

BACKGROUND: A large proportion of students identify statistics courses as the most anxiety-inducing courses in their curriculum. Many students feel impaired by feelings of state anxiety in the examination and therefore probably show lower achievements. AIMS: The study investigates how statistics anxiety, attitudes (e.g., interest, mathematical self-concept) and trait anxiety, as a general disposition to anxiety, influence experiences of anxiety as well as achievement in an examination. SAMPLE: Participants were 284 undergraduate psychology students, 225 females and 59 males. METHODS: Two weeks prior to the examination, participants completed a demographic questionnaire and measures of the STARS, the STAI, self-concept in mathematics, and interest in statistics. At the beginning of the statistics examination, students assessed their present state anxiety by the KUSTA scale. After 25 min, all examination participants gave another assessment of their anxiety at that moment. Students' examination scores were recorded. Structural equation modelling techniques were used to test relationships between the variables in a multivariate context. RESULTS: Statistics anxiety was the only variable related to state anxiety in the examination. Via state anxiety experienced before and during the examination, statistics anxiety had a negative influence on achievement. However, statistics anxiety also had a direct positive influence on achievement. This result may be explained by students' motivational goals in the specific educational setting. CONCLUSIONS: The results provide insight into the relationship between students' attitudes, dispositions, experiences of anxiety in the examination, and academic achievement, and give recommendations to instructors on how to support students prior to and in the examination.

Multilinear function factorisation for time series feature extraction

This work applies a variety of multilinear function factorisation techniques to extract appropriate features or attributes from high dimensional multivariate time series for classification. Recently, a great deal of work has centred around designing time series classifiers using more and more complex feature extraction and machine learning schemes. This paper argues that complex learners and domain specific feature extraction schemes of this type are not necessarily needed for time series classification, as excellent classification results can be obtained by simply applying a number of existing matrix factorisation or linear projection techniques, which are simple and computationally inexpensive. We highlight this using a geometric separability measure and classification accuracies obtained though experiments on four different high dimensional multivariate time series datasets. © 2013 IEEE.

Low-rank optimization with trace norm penalty

The paper addresses the problem of low-rank trace norm minimization. We propose an algorithm that alternates between fixed-rank optimization and rank-one updates. The fixed-rank optimization is characterized by an efficient factorization that makes the trace norm differentiable in the search space and the computation of duality gap numerically tractable. The search space is nonlinear but is equipped with a Riemannian structure that leads to efficient computations. We present a second-order trust-region algorithm with a guaranteed quadratic rate of convergence. Overall, the proposed optimization scheme converges superlinearly to the global solution while maintaining complexity that is linear in the number of rows and columns of the matrix. To compute a set of solutions efficiently for a grid of regularization parameters we propose a predictor-corrector approach that outperforms the naive warm-restart approach on the fixed-rank quotient manifold. The performance of the proposed algorithm is illustrated on problems of low-rank matrix completion and multivariate linear regression. © 2013 Society for Industrial and Applied Mathematics.