Dynamic load balancing in parallel KD-tree k-means

One among the most influential and popular data mining methods is the k-Means algorithm for cluster analysis. Techniques for improving the efficiency of k-Means have been largely explored in two main directions. The amount of computation can be significantly reduced by adopting geometrical constraints and an efficient data structure, notably a multidimensional binary search tree (KD-Tree). These techniques allow to reduce the number of distance computations the algorithm performs at each iteration. A second direction is parallel processing, where data and computation loads are distributed over many processing nodes. However, little work has been done to provide a parallel formulation of the efficient sequential techniques based on KD-Trees. Such approaches are expected to have an irregular distribution of computation load and can suffer from load imbalance. This issue has so far limited the adoption of these efficient k-Means variants in parallel computing environments. In this work, we provide a parallel formulation of the KD-Tree based k-Means algorithm for distributed memory systems and address its load balancing issue. Three solutions have been developed and tested. Two approaches are based on a static partitioning of the data set and a third solution incorporates a dynamic load balancing policy.

A hierarchical distributed approach for mining molecular fragments

Recently, two approaches have been introduced that distribute the molecular fragment mining problem. The first approach applies a master/worker topology, the second approach, a completely distributed peer-to-peer system, solves the scalability problem due to the bottleneck at the master node. However, in many real world scenarios the participating computing nodes cannot communicate directly due to administrative policies such as security restrictions. Thus, potential computing power is not accessible to accelerate the mining run. To solve this shortcoming, this work introduces a hierarchical topology of computing resources, which distributes the management over several levels and adapts to the natural structure of those multi-domain architectures. The most important aspect is the load balancing scheme, which has been designed and optimized for the hierarchical structure. The approach allows dynamic aggregation of heterogenous computing resources and is applied to wide area network scenarios.

A hierarchical Bayesian model for predicting the functional consequences of amino-acid polymorphisms

Genetic polymorphisms in deoxyribonucleic acid coding regions may have a phenotypic effect on the carrier, e.g. by influencing susceptibility to disease. Detection of deleterious mutations via association studies is hampered by the large number of candidate sites; therefore methods are needed to narrow down the search to the most promising sites. For this, a possible approach is to use structural and sequence-based information of the encoded protein to predict whether a mutation at a particular site is likely to disrupt the functionality of the protein itself. We propose a hierarchical Bayesian multivariate adaptive regression spline (BMARS) model for supervised learning in this context and assess its predictive performance by using data from mutagenesis experiments on lac repressor and lysozyme proteins. In these experiments, about 12 amino-acid substitutions were performed at each native amino-acid position and the effect on protein functionality was assessed. The training data thus consist of repeated observations at each position, which the hierarchical framework is needed to account for. The model is trained on the lac repressor data and tested on the lysozyme mutations and vice versa. In particular, we show that the hierarchical BMARS model, by allowing for the clustered nature of the data, yields lower out-of-sample misclassification rates compared with both a BMARS and a frequen-tist MARS model, a support vector machine classifier and an optimally pruned classification tree.

Synthesis of heterosulfamates. Search for structure-taste relationships

Thirty one new sodium heterosulfamates, RNHSO3Na, where the R portion contains mainly thiazole, benzothiazole, thiadiazole and pyridine ring structures, have been synthesized and their taste portfolios have been assessed. A database of 132 heterosulfamates ( both open-chain and cyclic) has been formed by combining these new compounds with an existing set of 101 heterosulfamates which were previously synthesized and for which taste data are available. Simple descriptors have been obtained using (i) measurements with Corey-Pauling-Koltun (CPK) space- filling models giving x, y and z dimensions and a volume VCPK, (ii) calculated first order molecular connectivities ((1)chi(v)) and (iii) the calculated Spartan program parameters to obtain HOMO, LUMO energies, the solvation energy E-solv and V-SPART AN. The techniques of linear (LDA) and quadratic (QDA) discriminant analysis and Tree analysis have then been employed to develop structure-taste relationships (SARs) that classify the sweet (S) and non-sweet (N) compounds into separate categories. In the LDA analysis 70% of the compounds were correctly classified ( this compares with 65% when the smaller data set of 101 compounds was used) and in the QDA analysis 68% were correctly classified ( compared to 80% previously). TheTree analysis correctly classified 81% ( compared to 86% previously). An alternative Tree analysis derived using the Cerius2 program and a set of physicochemical descriptors correctly classified only 54% of the compounds.

Modelling spectral reflectance of open cork oak woodland: a simulation analysis of the effects of vegetation structure and background

This study analyses the influence of vegetation structure (i.e. leaf area index and canopy cover) and seasonal background changes on moderate-resolution imaging spectrometer (MODIS)-simulated reflectance data in open woodland. Approximately monthly spectral reflectance and transmittance field measurements (May 2011 to October 2013) of cork oak tree leaves (Quercus suber) and of the herbaceous understorey were recorded in the region of Ribatejo, Portugal. The geometric-optical and radiative transfer (GORT) model was used to simulate MODIS response (red, near-infrared) and to calculate vegetation indices, investigating their response to changes in the structure of the overstorey vegetation and to seasonal changes in the understorey using scenarios corresponding to contrasting phenological status (dry season vs. wet season). The performance of normalized difference vegetation index (NDVI), soil-adjusted vegetation index (SAVI), and enhanced vegetation index (EVI) is discussed. Results showed that SAVI and EVI were very sensitive to the emergence of background vegetation in the wet season compared to NDVI and that shading effects lead to an opposing trend in the vegetation indices. The information provided by this research can be useful to improve our understanding of the temporal dynamic of vegetation, monitored by vegetation indices.

THE ENDOPARASITE PILOSTYLES ULEI (APODANTHACEAE - CUCURBITALES) INFLUENCES WOOD STRUCTURE IN THREE HOST SPECIES OF MIMOSA

Pilostyles species (Apodanthaceae) are endoparasites in stems of the plant family Fabaceae. The body comprises masses of parenchyma in the host bark and cortex, with sinkers, comprising groups of twisted tracheal elements surrounded by parenchyma that enter the secondary xylem of the host plant. Here we report for the first time the effects of Pilostyles parasitism on host secondary xylem. We obtained healthy and parasitized stems from Mimosa foliolosa, M. maguirei and M. setosa and compared vessel element length, fiber length, vessel diameter and vessel frequency, measured through digital imaging. Also, tree height and girth were compared between healthy and parasitized M. setosa. When parasitized, plant size, vessel diameter, vessel element length and fiber length are all less than in healthy plants. Also, vessel frequency is greater and vessels are narrower in parasitized stems. These responses to parasitism are similar to those observed in stressed plants. Thus, hosts respond to the parasite by changing its wood micromorphology in favour of increased hydraulic safety.

Cell wall polysaccharides from cell suspension cultures of the Atlantic Forest tree Rudgea jasminoides (Rubiaceae)

Rudgea jasminoides (Rubiaceae) is a tropical tree species native of the Atlantic Forest in the south of Brazil. Previous studies with leaf cell walls of R. jasminoides showed a different proportion of cross-linked glycans compared to what is usually reported for eudicots. However, due to the difficulties of working with whole plant organs, cell suspensions of R. jasminoides, consisting of predominantly undifferentiated cells with mainly primary cell walls, were used to examine cell walls and extracellular soluble polysaccharides (EP) released into the culture medium. Sugar composition and linkage analysis showed homogalacturonans, xylogalacturonans and arabinogalactans to be the predominant EP. In the cell wall, homogalacturonans and arabinogalactans are the major pectins, and xyloglucans and xylans are the major cross-linking glycans. The presence of xylogalacturonans in the R. jasminoides cell cultures seems to be related to the occurrence of a homogeneous cell suspension with loosely attached cells. Although all alkali extractions from the cell walls yielded amounts of xyloglucan that exceed those of the xylans, the latter was found in a proportion that is higher than what has been usually reported for primary cell walls of most eudicots. The xyloglucan from cell walls of cell suspension cultures of R. jasminoides has low fucosylation levels and high proportion of galactosyl residues, a branching pattern commonly found in storage cell-wall xyloglucans.

Forest structure in a mosaic of rainforest sites: The effect of fragmentation and recovery after clear cut

A variety of human-induced disturbances such as forest fragmentation and recovery after deforestation for pasture or agricultural activities have resulted in a complex landscape mosaic in the Una region of northeastern Brazil. Using a set of vegetation descriptors, we investigated the main structural changes observed in forest categories that comprise the major components of the regional landscape and searched for potential key descriptors that could be used to discriminate among different forest categories. We assessed the forest structure of five habitat categories defined as (I) interiors and (2) edges of large fragments of old-growth forest (>1000 ha), (3) interiors and (4) edges of small forest fragments (<100 ha), and (5) early secondary forests. Forest descriptors used here were: frequency of herbaceous lianas and woody climbers, number of standing dead trees, number of fallen trunks, litter depth, number of pioneer plants (early secondary and shade-intolerant species), vertical foliage stratification profile and distribution Of trees in different diameter classes. Edges and interiors of forest fragments were significantly different only in the number of standing dead trees. Secondary forests and edges of fragments showed differences in litter depth, fallen trunks and number of pioneer trees, and secondary forests were significantly different from fragment interiors in the number of standing dead trees and the number of pioneer trees. Horizontal and vertical structure evaluated via ordination analysis showed that fragment interiors, compared to secondary forests, were characterized by a greater number of medium (25-35 cm) and large (35-50 cm) trees and smaller numbers of thin trees (5-10 cm). There was great heterogeneity at the edges of small and large fragments, as these sites were distributed along almost the entire gradient. Most interiors of large and small fragments presented higher values of foliage densities at higher strata ( 15-20 m and at 20-25 m height), and lower densities at 1-5 m. All secondary forests and some fragment edge sites showed an opposite tendency. A discriminant function highlighted differences among forest categories, with transects of large fragment interiors and secondary forests representing two extremes along a disturbance gradient determined by foliage structure (densities at 15-20 m and 20-25 m), with the edges of both large and small fragments and the interiors of small fragments scattered across the gradient. The major underlying processes determining patterns of forest disturbance in the study region are discussed, highlighting the importance of forest fragments, independently of its size, as forests recovery after clear cut show a greatly distinct structure, with profound implications on fauna movements. (C) 2009 Elsevier BY. All rights reserved.

Building binary-tree-based multiclass classifiers using separability measures

Various popular machine learning techniques, like support vector machines, are originally conceived for the solution of two-class (binary) classification problems. However, a large number of real problems present more than two classes. A common approach to generalize binary learning techniques to solve problems with more than two classes, also known as multiclass classification problems, consists of hierarchically decomposing the multiclass problem into multiple binary sub-problems, whose outputs are combined to define the predicted class. This strategy results in a tree of binary classifiers, where each internal node corresponds to a binary classifier distinguishing two groups of classes and the leaf nodes correspond to the problem classes. This paper investigates how measures of the separability between classes can be employed in the construction of binary-tree-based multiclass classifiers, adapting the decompositions performed to each particular multiclass problem. (C) 2010 Elsevier B.V. All rights reserved.

HiPP: A Novel Hierarchical Point Placement Strategy and its Application to the Exploration of Document Collections

Point placement strategies aim at mapping data points represented in higher dimensions to bi-dimensional spaces and are frequently used to visualize relationships amongst data instances. They have been valuable tools for analysis and exploration of data sets of various kinds. Many conventional techniques, however, do not behave well when the number of dimensions is high, such as in the case of documents collections. Later approaches handle that shortcoming, but may cause too much clutter to allow flexible exploration to take place. In this work we present a novel hierarchical point placement technique that is capable of dealing with these problems. While good grouping and separation of data with high similarity is maintained without increasing computation cost, its hierarchical structure lends itself both to exploration in various levels of detail and to handling data in subsets, improving analysis capability and also allowing manipulation of larger data sets.

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.

A Tree Tools informática e o processo de implantação da mudança administrativa de formação do tipo de capital e suas conseqüências nas áreas motivacionais e de satisfação

A organização do trabalho constitui-se a partir da caracterização de sua estrutura e dos respectivos impactos do desempenho de papéis profissionais. O relacionamento interpessoal no ambiente de trabalho reflete o grau de participação e colaboração dos indivíduos. Os indivíduos diferem uns dos outros pelas suas atitudes em relação a um trabalho "estimulante" ou não. O valor prático do estudo, busca analisar o grau de satisfação I insatisfação dos indivíduos em relação a si próprios e em relação a empresa. Também foi feito um estudo das abordagens motivacionais e suas ligações com os indivíduos acima citados. A maior parte dos autores que contribuíram ao desenvolvimento dos temas considerados neste estudo parecem ter querido conciliar dois imperativos: aumentar a rendimento dos trabalhadores e elevar o seu nível de satisfação no trabalho (este segundo imperativo foi reformulado recentemente em termos de melhora da qualidade de vida no trabalho). Ainda no corpo do trabalho foram apresentados resultados da pesquisa realizada e discutidas algumas das abordagens motivacionais, com a finalidade de fundamentar teoricamente a questão principal que é a forma de administração da empresa o que se trata de satisfação I motivação na empresa TREE TOOLS, escolhida como agente deste estudo de caso.

Processing of hierarchical syntactic strusture in music

Hierarchical structure with nested nonlocal dependencies is a key feature of human language and can be identified theoretically in most pieces of tonal music. However, previous studies have argued against the perception of such structures in music. Here, we show processing of nonlocal dependencies in music. We presented chorales by J. S. Bach and modified versions inwhich the hierarchical structure was rendered irregular whereas the local structure was kept intact. Brain electric responses differed between regular and irregular hierarchical structures, in both musicians and nonmusicians. This finding indicates that, when listening to music, humans apply cognitive processes that are capable of dealing with longdistance dependencies resulting from hierarchically organized syntactic structures. Our results reveal that a brain mechanism fundamental for syntactic processing is engaged during the perception of music, indicating that processing of hierarchical structure with nested nonlocal dependencies is not just a key component of human language, but a multidomain capacity of human cognition.

Dispersion of Seeds of Tree Species by the Leaf-Cutting Ant Acromyrmex subterraneus molestans (Hymenoptera: Formicidae) in Vicosa, Minas Gerais State, Brazil

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

An evergreen neotropical savanna tree (Gochnatia polymorpha, Asteraceae) produces different dry- and wet-season leaf types

Gochnatia polymorpha (Less.) Cabrera is a widespread tree species found in different physiognomies of neotropical savanna (cerrado) formations of south-eastern Brazil. The present study describes some leaf anatomical characteristics of this species as a function of the time of leaf flush, during dry or wet seasons. This species presents anatomical plasticity in the cuticle, palisade parenchyma and abaxial epidermis as well as in stomatal size and stomatal and trichome density, which are leaf structures linked with water-status control. Leaf structure changed to suit the particular environmental conditions during dry and wet seasons. The production of different wet-and dry-season leaf types in G. polymorpha could be a response to drought and an adaptation to environmental constraints in the cerrado.