HIERARCHICAL DECOMPOSITION OF MULTICLASS PROBLEMS

Several popular Machine Learning techniques are originally designed for the solution of two-class problems. However, several classification problems have more than two classes. One approach to deal with multiclass problems using binary classifiers is to decompose the multiclass problem into multiple binary sub-problems disposed in a binary tree. This approach requires a binary partition of the classes for each node of the tree, which defines the tree structure. This paper presents two algorithms to determine the tree structure taking into account information collected from the used dataset. This approach allows the tree structure to be determined automatically for any multiclass dataset.

Hierarchical spherical model from a geometric point of view

A continuous version of the hierarchical spherical model at dimension d=4 is investigated. Two limit distributions of the block spin variable X(gamma), normalized with exponents gamma = d + 2 and gamma=d at and above the critical temperature, are established. These results are proven by solving certain evolution equations corresponding to the renormalization group (RG) transformation of the O(N) hierarchical spin model of block size L(d) in the limit L down arrow 1 and N ->infinity. Starting far away from the stationary Gaussian fixed point the trajectories of these dynamical system pass through two different regimes with distinguishable crossover behavior. An interpretation of this trajectories is given by the geometric theory of functions which describe precisely the motion of the Lee-Yang zeroes. The large-N limit of RG transformation with L(d) fixed equal to 2, at the criticality, has recently been investigated in both weak and strong (coupling) regimes by Watanabe (J. Stat. Phys. 115:1669-1713, 2004) . Although our analysis deals only with N = infinity case, it complements various aspects of that work.

Hierarchical spatial organization of geographical networks

In this work, we propose a hierarchical extension of the polygonality index as the means to characterize geographical planar networks. By considering successive neighborhoods around each node, it is possible to obtain more complete information about the spatial order of the network at progressive spatial scales. The potential of the methodology is illustrated with respect to synthetic and real geographical networks.

Hierarchical Assembly of CaMoO(4) Nano-Octahedrons and Their Photoluminescence Properties

Hierarchical assemblies of CaMoO4 (CM) nano-octahedrons were obtained by microwave-assisted hydrothemial synthesis at 120 degrees C for different times. These structures were structurally, morphologically and optically characterized by X-ray diffraction, micro-Raman spectroscopy, field-emission gun scanning electron microscopy, ultraviolet-visible absorption spectroscopy and photoluminescence measurements. First-principle calculations have been carried out to understand the structural and electronic order-disorder effects as a function of the particle/region size. Supercells of different dimensions were constructed to simulate the geometric distortions along both they and z planes of the scheelite structure. Based on these experimental results and with the help of detailed structural simulations, we were able to model the nature of the order-disorder in this important class of materials and discuss the consequent implications on its physical properties, in particular, the photoluminescence properties of CM nanocrystals.

Estimation of genetic variance for macro- and micro-environmental sensitivity using double hierarchical generalized linear models

Background: Genetic variation for environmental sensitivity indicates that animals are genetically different in their response to environmental factors. Environmental factors are either identifiable (e.g. temperature) and called macro-environmental or unknown and called micro-environmental. The objectives of this study were to develop a statistical method to estimate genetic parameters for macro- and micro-environmental sensitivities simultaneously, to investigate bias and precision of resulting estimates of genetic parameters and to develop and evaluate use of Akaike’s information criterion using h-likelihood to select the best fitting model. Methods: We assumed that genetic variation in macro- and micro-environmental sensitivities is expressed as genetic variance in the slope of a linear reaction norm and environmental variance, respectively. A reaction norm model to estimate genetic variance for macro-environmental sensitivity was combined with a structural model for residual variance to estimate genetic variance for micro-environmental sensitivity using a double hierarchical generalized linear model in ASReml. Akaike’s information criterion was constructed as model selection criterion using approximated h-likelihood. Populations of sires with large half-sib offspring groups were simulated to investigate bias and precision of estimated genetic parameters. Results: Designs with 100 sires, each with at least 100 offspring, are required to have standard deviations of estimated variances lower than 50% of the true value. When the number of offspring increased, standard deviations of estimates across replicates decreased substantially, especially for genetic variances of macro- and micro-environmental sensitivities. Standard deviations of estimated genetic correlations across replicates were quite large (between 0.1 and 0.4), especially when sires had few offspring. Practically, no bias was observed for estimates of any of the parameters. Using Akaike’s information criterion the true genetic model was selected as the best statistical model in at least 90% of 100 replicates when the number of offspring per sire was 100. Application of the model to lactation milk yield in dairy cattle showed that genetic variance for micro- and macro-environmental sensitivities existed. Conclusion: The algorithm and model selection criterion presented here can contribute to better understand genetic control of macro- and micro-environmental sensitivities. Designs or datasets should have at least 100 sires each with 100 offspring.

Hglm: A package for fitting hierarchical generalized linear models

We present the hglm package for fitting hierarchical generalized linear models. It can be used for linear mixed models and generalized linear mixed models with random effects for a variety of links and a variety of distributions for both the outcomes and the random effects. Fixed effects can also be fitted in the dispersion part of the model.

Genetic heterogeneity of residual variance : estimation of variance components using double hierarchical generalized linear models

Background: The sensitivity to microenvironmental changes varies among animals and may be under genetic control. It is essential to take this element into account when aiming at breeding robust farm animals. Here, linear mixed models with genetic effects in the residual variance part of the model can be used. Such models have previously been fitted using EM and MCMC algorithms. Results: We propose the use of double hierarchical generalized linear models (DHGLM), where the squared residuals are assumed to be gamma distributed and the residual variance is fitted using a generalized linear model. The algorithm iterates between two sets of mixed model equations, one on the level of observations and one on the level of variances. The method was validated using simulations and also by re-analyzing a data set on pig litter size that was previously analyzed using a Bayesian approach. The pig litter size data contained 10,060 records from 4,149 sows. The DHGLM was implemented using the ASReml software and the algorithm converged within three minutes on a Linux server. The estimates were similar to those previously obtained using Bayesian methodology, especially the variance components in the residual variance part of the model. Conclusions: We have shown that variance components in the residual variance part of a linear mixed model can be estimated using a DHGLM approach. The method enables analyses of animal models with large numbers of observations. An important future development of the DHGLM methodology is to include the genetic correlation between the random effects in the mean and residual variance parts of the model as a parameter of the DHGLM.

Automatically Adapting Source Code to Document Provenance

Being able to ask questions about the provenance of some data requires documentation on each influence on that data's existence and content. Much software exists, and is being developed, for which there is no provenance-awareness, i.e. at best, the data it outputs can be connected to its inputs, but with no record of intermediate processing. Further, where some record of processing does exist, e.g. as logs, it is not in a form easily connected with that of other processes. We would like to enable compiled software to record useful documentation without requiring prior manual adaptation. In this paper, we present an approach to adapting source code from its original form without manual manipulation, to record information on data provenance during execution.

Measuring Australia's well-being using hierarchical needs

A new approach to well-being measurement is presented in this paper based on multidimensional hierarchical human needs and motivation. This paper empirically applies this new measure of well-being to Australia for the period 1985–2000. This hierarchical approach is underpinned by a rigorous psychological theory of human motivation. Hierarchical human needs are classified into five categories. Eight indicators have been chosen to reflect these categories. A composite indicator of these eight indicators is calculated. This paper concludes that it is necessary to consider multidimensional human needs and motivation when analysing and seeking to improve well-being through economic and social development activities.

Phase behavior, crystallization, and hierarchical nanostructures in self-organized thermoset blends of epoxy resin and amphiphilic poly(ethylene oxide)-block-poly(propylene oxide)-block-poly(ethylene oxide) triblock copolymers

Nanostructured thermoset blends of bisphenol A-type epoxy resin (ER) and amphiphilic poly(ethylene oxide)-block-poly(propylene oxide)-block-poly(ethylene oxide) (PEO-PPO-PEO) triblock copolymers were successfully prepared. Two samples of PEO-PPO-PEO triblock copolymer with different ethylene oxide (EO) contents, denoted as EO30 with 30 wt % EO content and EO80 with 80 wt % EO content, were used to form the self-organized thermoset blends of varying compositions using 4,4'-methylenedianiline (MDA) as curing agent. The phase behavior, crystallization, and morphology were investigated by differential scanning calorimetry (DSC), transmission electron microscopy (TEM), atomic force microscopy (AFM), and small-angle X-ray scattering (SAXS). It was found that macroscopic phase separation took place in the MDA-cured ER/EO30 blends containing 60-80 wt % EO30 triblock copolymer. The MDA-cured ER/EO30 blends with EO30 content up to 50 wt % do not show macroscopic phase separation but exhibit nanostructures on the order of 10-30 nm as revealed by both the TEM and SAXS studies. The AFM study further shows that the ER/EO30 blend at some composition displays structural inhomogeneity at two different nanoscales and is hierarchically nanostructured. The spherical PPO domains with an average size of about 10 nm are uniformly dispersed in the 80/20 ER/EO30 blend; meanwhile, a structural inhomogeneity on the order of 50-200 nm is observed. The ER/EO80 blends are not macroscopically phase-separated over the entire composition range because of the much higher PEO content of the EO80 triblock copolymer. However, the ER/EO80 blends show composition-dependent nanostructures on the order of 10-100 nm. The 80/20 ER/EO80 blend displays hierarchical structures at two different nanoscales, i.e., a bicontinuous microphase structure on the order of about 100 nm and spherical domains of 10-20 nm in diameter uniformly dispersed in both the continuous microphases. The blends with 60 wt % and higher EO80 content are completely volume-filled with spherulites. Bundles of PEO lamellae with spacing of 20-30 nm interwoven with a microphase structure on the order of about 100 nm are revealed by AFM study for the 30/70 ER/EO80 blend.

Web page clustering : a hyperlink-based similarity and matrix-based hierarchical algorithms

This paper proposes a hyperlink-based web page similarity measurement and two matrix-based hierarchical web page clustering algorithms. The web page similarity measurement incorporates hyperlink transitivity and page importance within the concerned web page space. One clustering algorithm takes cluster overlapping into account, another one does not. These algorithxms do not require predefined similarity thresholds for clustering, and are independent of the page order. The primary evaluations show the effectiveness of the proposed algorithms in clustering improvement.

Interpersonal trust and formal controls in hierarchical relationships

This paper examines the influence of the level of interpersonal trust between superior and subordinate managers on the control behaviour of the former. On the basis of a questionnaire survey and interviews of senior managers from business organisations in Sri Lanka, and a survey of managers in Beijing-China the study explores the control behaviour of superior managers when their trust in a subordinate is high or low. Sri Lanka and China, societies in which the dependence on interpersonal trust is believed to be high, were chosen for the study to maximise the effect of interpersonal trust.

The findings of this study indicate that a superior’s high (low) trust in a subordinate is associated with a low (high) level of monitoring and a high (low) level of social interactions. The hypothesis that a superior’s high (low) level of trust is associated with a low (high) level of reliance on formal control was supported only in the Sri Lankan sample. These findings are at least indicative of control behaviour of superior managers in Sri Lanka and China and possibly of other countries in Asia. An understanding of the trust-sensitive control behaviour of managers in this region is particularly important for designing and implementing effective control systems for international organizations operating in the region.

A maximal frequent itemset approach for web document clustering

To efficiently and yet accurately cluster Web documents is of great interests to Web users and is a key component of the searching accuracy of a Web search engine. To achieve this, this paper introduces a new approach for the clustering of Web documents, which is called maximal frequent itemset (MFI) approach. Iterative clustering algorithms, such as K-means and expectation-maximization (EM), are sensitive to their initial conditions. MFI approach firstly locates the center points of high density clusters precisely. These center points then are used as initial points for the K-means algorithm. Our experimental results tested on 3 Web document sets show that our MFI approach outperforms the other methods we compared in most cases, particularly in the case of large number of categories in Web document sets.