Hierarchical Mixtures of Experts and the EM Algorithm

We present a tree-structured architecture for supervised learning. The statistical model underlying the architecture is a hierarchical mixture model in which both the mixture coefficients and the mixture components are generalized linear models (GLIM's). Learning is treated as a maximum likelihood problem; in particular, we present an Expectation-Maximization (EM) algorithm for adjusting the parameters of the architecture. We also develop an on-line learning algorithm in which the parameters are updated incrementally. Comparative simulation results are presented in the robot dynamics domain.

Selecting Relevant Genes with a Spectral Approach

Array technologies have made it possible to record simultaneously the expression pattern of thousands of genes. A fundamental problem in the analysis of gene expression data is the identification of highly relevant genes that either discriminate between phenotypic labels or are important with respect to the cellular process studied in the experiment: for example cell cycle or heat shock in yeast experiments, chemical or genetic perturbations of mammalian cell lines, and genes involved in class discovery for human tumors. In this paper we focus on the task of unsupervised gene selection. The problem of selecting a small subset of genes is particularly challenging as the datasets involved are typically characterized by a very small sample size ?? the order of few tens of tissue samples ??d by a very large feature space as the number of genes tend to be in the high thousands. We propose a model independent approach which scores candidate gene selections using spectral properties of the candidate affinity matrix. The algorithm is very straightforward to implement yet contains a number of remarkable properties which guarantee consistent sparse selections. To illustrate the value of our approach we applied our algorithm on five different datasets. The first consists of time course data from four well studied Hematopoietic cell lines (HL-60, Jurkat, NB4, and U937). The other four datasets include three well studied treatment outcomes (large cell lymphoma, childhood medulloblastomas, breast tumors) and one unpublished dataset (lymph status). We compared our approach both with other unsupervised methods (SOM,PCA,GS) and with supervised methods (SNR,RMB,RFE). The results clearly show that our approach considerably outperforms all the other unsupervised approaches in our study, is competitive with supervised methods and in some case even outperforms supervised approaches.

Validating Co-Training Models for Web Image Classification

Co-training is a semi-supervised learning method that is designed to take advantage of the redundancy that is present when the object to be identified has multiple descriptions. Co-training is known to work well when the multiple descriptions are conditional independent given the class of the object. The presence of multiple descriptions of objects in the form of text, images, audio and video in multimedia applications appears to provide redundancy in the form that may be suitable for co-training. In this paper, we investigate the suitability of utilizing text and image data from the Web for co-training. We perform measurements to find indications of conditional independence in the texts and images obtained from the Web. Our measurements suggest that conditional independence is likely to be present in the data. Our experiments, within a relevance feedback framework to test whether a method that exploits the conditional independence outperforms methods that do not, also indicate that better performance can indeed be obtained by designing algorithms that exploit this form of the redundancy when it is present.

Scene Classification Using a Hybrid Generative/Discriminative Approach

We investigate whether dimensionality reduction using a latent generative model is beneficial for the task of weakly supervised scene classification. In detail, we are given a set of labeled images of scenes (for example, coast, forest, city, river, etc.), and our objective is to classify a new image into one of these categories. Our approach consists of first discovering latent ";topics"; using probabilistic Latent Semantic Analysis (pLSA), a generative model from the statistical text literature here applied to a bag of visual words representation for each image, and subsequently, training a multiway classifier on the topic distribution vector for each image. We compare this approach to that of representing each image by a bag of visual words vector directly and training a multiway classifier on these vectors. To this end, we introduce a novel vocabulary using dense color SIFT descriptors and then investigate the classification performance under changes in the size of the visual vocabulary, the number of latent topics learned, and the type of discriminative classifier used (k-nearest neighbor or SVM). We achieve superior classification performance to recent publications that have used a bag of visual word representation, in all cases, using the authors' own data sets and testing protocols. We also investigate the gain in adding spatial information. We show applications to image retrieval with relevance feedback and to scene classification in videos

Implementation of ‘School Back’ by the ‘Core’ technique in a factory of geotextiles to prevent back pain

Introduction: This article aims to show an alternative intervention for the prevention and control of back pain to the people of a production plant of geotextiles for the construction exposed to handling and awkward postures through the implementation of the Back School using the CORE technique. This technique being understood as trainer of the stability musculature of the spine; whose benefit is proportionate the muscular complex of the back, stability and avoid osteomuscular lesions and improved posture. Objective: To present the results about the implementation of the back school by the CORE technique for prevention of back pain in a population of forty-eight male collaborators. Materials and methods: The back school began with talks of awareness by the occupational health physician explaining the objectives and benefits of it to all participants. Once this activity was done, was continued to evaluate all plant employees to establish health status through the PAR-Q questionnaire, who were surveyed for the perception of pain using visual analog scale (VAS) and stability was determined column through the CORE assessment, to determine the training plan. Then, were made every six months the revaluations and implementation of a survey of assistant public perception to identify the impact of the implementation of the school back on the two variables referred (pain perception and stability of column). Results: The pain perception according VAS increased in the number of workers asymptomatic in 12% and based in the satisfaction survey 94% of population reported that with the development of this technique decrease the muscle fatigue in lumbar level; and 96% of population reported an improvement in the performance of their work activities. Discussion: Posterior to the analysis of all results, it is interpreted that back schools practice through CORE technique, contributes to the prevention and / or control of symptoms at the lumbar level in population of productive sector exposed to risks derived from the physical load, provided that ensure its continuously development and supervised for a competent professional.

Revisión sistemática: programas de ejercicio en rehabilitación cardiaca intrahospitalaria versus rehabilitación cardiaca en casa para pacientes con evento coronario agudo

El objetivo de la rehabilitación cardiaca es lograr que los pacientes con enfermedades cardiacas, reanuden su vida activa y productiva; logrando un óptimo estado físico, psicosocial y vocacional. Los programas de ejercicio, son parte básica de la rehabilitación cardiaca. Existen muchos tipos de programas de actividad física, que varían entre programas de ejercicio en casa no supervisados, hasta programas intrahospitalarios altamente supervisados.

Factores de riesgo del comportamiento y readiness de aficionados inscritos a carreras atléticas de fondo de 18-64 años en Bogotá D.C., 2014”

La participación en carreras atléticas de calle ha aumentado; esto requiere detectar riesgos previos al esfuerzo físico. Objetivo. Identificar factores de riesgo del comportamiento y readiness de inscritos a una carrera. Método. Estudio transversal en aficionados de 18-64 años. Encuesta digital con módulos de IPAQ, PARQ+ y STEP. Muestreo aleatorio sistemático con n=510, para una inactividad física esperada de 35% (±5%). Se evaluó nivel de actividad física, consumo de alcohol (peligroso), de fruta, verdura, tabaco y sal, y readiness. Resultados. El cumplimiento de actividad física fue 97,4%; 2,4% consume nivel óptimo de fruta o verdura (diferencias por edad, sexo y estrato), 3,7% fuma y 44,1% consumo peligroso de alcohol. El 19,8% reportó PARQ+ positivo y 5,7% requiere supervisión. Hay diferencias por trabajo y estudio. Discusión. Los aficionados cumplen el nivel de actividad física; pero no de otros factores. Una estrategia de seguridad en el atletismo de calle es evaluar los factores de riesgo relacionados con el estilo de vida así como el readiness.

Image classification for a large number of object categories

L'increment de bases de dades que cada vegada contenen imatges més difícils i amb un nombre més elevat de categories, està forçant el desenvolupament de tècniques de representació d'imatges que siguin discriminatives quan es vol treballar amb múltiples classes i d'algorismes que siguin eficients en l'aprenentatge i classificació. Aquesta tesi explora el problema de classificar les imatges segons l'objecte que contenen quan es disposa d'un gran nombre de categories. Primerament s'investiga com un sistema híbrid format per un model generatiu i un model discriminatiu pot beneficiar la tasca de classificació d'imatges on el nivell d'anotació humà sigui mínim. Per aquesta tasca introduïm un nou vocabulari utilitzant una representació densa de descriptors color-SIFT, i desprès s'investiga com els diferents paràmetres afecten la classificació final. Tot seguit es proposa un mètode par tal d'incorporar informació espacial amb el sistema híbrid, mostrant que la informació de context es de gran ajuda per la classificació d'imatges. Desprès introduïm un nou descriptor de forma que representa la imatge segons la seva forma local i la seva forma espacial, tot junt amb un kernel que incorpora aquesta informació espacial en forma piramidal. La forma es representada per un vector compacte obtenint un descriptor molt adequat per ésser utilitzat amb algorismes d'aprenentatge amb kernels. Els experiments realitzats postren que aquesta informació de forma te uns resultats semblants (i a vegades millors) als descriptors basats en aparença. També s'investiga com diferents característiques es poden combinar per ésser utilitzades en la classificació d'imatges i es mostra com el descriptor de forma proposat juntament amb un descriptor d'aparença millora substancialment la classificació. Finalment es descriu un algoritme que detecta les regions d'interès automàticament durant l'entrenament i la classificació. Això proporciona un mètode per inhibir el fons de la imatge i afegeix invariança a la posició dels objectes dins les imatges. S'ensenya que la forma i l'aparença sobre aquesta regió d'interès i utilitzant els classificadors random forests millora la classificació i el temps computacional. Es comparen els postres resultats amb resultats de la literatura utilitzant les mateixes bases de dades que els autors Aixa com els mateixos protocols d'aprenentatge i classificació. Es veu com totes les innovacions introduïdes incrementen la classificació final de les imatges.

O Supervisor como mediador no processo do desenvolvimento pessoal e profissional do professor : do ideal ao real na escola do séc. XXI em Portugal

O presente estudo de investigação visa percepcionar o papel e o trabalho de um supervisor pedagógico na escola portuguesa que possibilite compreender os equívocos e deturpação ligados ao papel do supervisor, quer por quem exerce a função quer por aqueles que são supervisionados. A revisão da literatura foi orientada no sentido de definir a acção e competências do supervisor e os efeitos no desenvolvimento pessoal e profissional do professor avaliado num contexto de uma escola aprendente ou reflexiva, segundo algumas abordagens mais modernas, que gera um trabalho colaborativo entre pares e leva a melhoria do processo ensino-aprendizagem. Analisaram-se desta forma quais os aspectos mais relevantes para nomeação do supervisor, competências e trabalho a desenvolver por este e efeitos no desenvolvimento pessoal e profissional do professor supervisionado e ainda as dificuldades sentidas na concretização da supervisão pedagógica. A investigação decorreu numa Escola do concelho de Matosinhos, recorreu-se a procedimentos metodológicos quantitativos através de inquéritos por questionário aplicados aos professores Relatores/Supervisores e aos professores Avaliados/Supervisionados. Desta investigação ressalta um desfasamento entre os subgrupos inquiridos, sobretudo em relação aos aspectos a ter em conta para sua nomeação, perfil e competências inerentes à função, e no trabalho a desenvolver e desenvolvido pelo Relator/Supervisor.

O efeito de um programa de exercício físico na capacidade funcional, nos fatores de risco de progressão da doença coronária e na qualidade de vida em doentes cardíacos

Considerando que a prática de exercício físico com uma intensidade pelo menos moderada melhora a capacidade funcional (Maines et al., 1997; Clara et al., 2002; Olney et al., 2006), a qualidade de vida (Leal et al., 2005; Azevedo & Leal, 2009; Flynn et al., 2009) e diminui os fatores de risco coronários (Maines et al., 1997; Squires & Hamm, 2007; Perk, 2009; Pimenta, 2010), propõe-se com o presente estudo analisar o efeito do exercício físico supervisionado, em fase ambulatório precoce, realizada na comunidade, ao nível da recuperação de doentes cardíacos. Método: Aplicar-se-á um estudo experimental, em doentes cardíacos de ambos os sexos, entre os 28 e os 80 anos. Atribuir-se-á particular ênfase às alterações induzidas pela aplicação do programa de exercício físico nos parâmetros bioquímicos (colesterol total, C-LDL, C-HDL, triglicéridos e glicose), na composição corporal (peso, índice de massa corporal, perímetro da cintura), na capacidade funcional (consumo de oxigénio pico –V02 pico, equivalente metabólico, duplo produto), no nível de atividade física, na ingestão alimentar e na qualidade de vida. O estudo terá uma duração superior a três meses, comparando dois grupos, um grupo submetido ao exercício físico supervisionado (ES) e outro aos cuidados usuais (CU), os quais serão alvo de duas avaliações (inicial e final), avaliando-se a média e o desvio-padrão para todas as variáveis em estudo e recorrendo-se aos testes não paramétricos e paramétricos, para um nível de significância de p< .05. Resultados: Foram elegidos 52 doentes, sendo que 22 participaram no grupo cuidados usuais (CU) e 30 no grupo exercício físico supervisionado (ES), observando-se que o grupo ES apresentou melhorias mais acentuadas quando comparadas com o grupo CU, ao nível dos seguintes indicadores: dispêndio de kcal/semana (+697.22% vs +320.20%); PC (-3.19% vs +5.85%); CT (-23.92% vs -9.29%), C-LDL (-32.52% vs -8.92%); total de kcal/dia ingeridas (-33,31% vs -2.58%); VO2 pico (+30.88% vs -3.57%); qualidade de vida geral (+53.86% vs +2.96%). Conclusão: Concluindo que o exercício físico multicomponente, inserido na fase de ambulatório precoce na comunidade, potencia a recuperação de doentes cardíacos influenciando positivamente os fatores de risco de progressão da doença coronária, a capacidade funcional e a qualidade de vida fundamentais para que o doente possa, pelos seus próprios meios, retomar a sua vida na comunidade.

O portefólio como instrumento de reflexão na disciplina de desenho A do Ensino Secundário

A temática apresentada é resultante da prática de ensino supervisionada no Ensino Secundário em Artes Visuais e tem como finalidade aprofundar questões que envolvem a autorregulação das aprendizagens para o desenvolvimento do pensamento reflexivo dos alunos. A importância do paradigma reflexivo está cada vez mais presente nas práticas educativas dos nossos tempos, remetendo para um novo conceito de ensino e de avaliação no processo de ensino-aprendizagem, de acordo com um ensino cada vez mais centrado no aluno. Seguindo a intemporalidade do pensamento freireano são desenvolvidas práticas de avaliação tendo por base a integração do portefólio como instrumento de aprendizagem progressiva para a prática reflexiva dos alunos na disciplina de Desenho A. Este projeto de investigação participativa pretende analisar informações qualitativas e quantitativas de forma a reconhecer o contributo e o benefício da dimensão reflexiva ao longo do processo de ensino-aprendizagem no desenvolvimento pessoal e cognitivo do aluno.

Evaluating the Potential Impact of a Gas Pipeline on Whimbrel Breeding Habitat in the Outer Mackenzie Delta, Northwest Territories

We used ground surveys to identify breeding habitat for Whimbrel (Numenius phaeopus) in the outer Mackenzie Delta, Northwest Territories, and to test the value of high-resolution IKONOS imagery for mapping additional breeding habitat in the Delta. During ground surveys, we found Whimbrel nests (n = 28) in extensive areas of wet-sedge low-centered polygon (LCP) habitat on two islands in the Delta (Taglu and Fish islands) in 2006 and 2007. Supervised classification using spectral analysis of IKONOS imagery successfully identified additional areas of wet-sedge habitat in the region. However, ground surveys to test this classification found that many areas of wet-sedge habitat had dense shrubs, no standing water, and/or lacked polygon structure and did not support breeding Whimbrel. Visual examination of the IKONOS imagery was necessary to determine which areas exhibited LCP structure. Much lower densities of nesting Whimbrel were also found in upland habitats near wetlands. We used habitat maps developed from a combination of methods, to perform scenario analyses to estimate the potential effects of the Mackenzie Gas Project on Whimbrel habitat. Assuming effective complete habitat loss within 20 m, 50 m, or 250 m of any infrastructure or pipeline, the currently proposed pipeline development would result in loss of 8%, 12%, or 30% of existing Whimbrel habitat. If subsidence were to occur, most Whimbrel habitat could become unsuitable. If the facility is developed, follow-up surveys will be required to test these models.

Detection of flood extent in urban areas using high resolution TerraSAR-X data

Flooding is a major hazard in both rural and urban areas worldwide, but it is in urban areas that the impacts are most severe. An investigation of the ability of high resolution TerraSAR-X data to detect flooded regions in urban areas is described. An important application for this would be the calibration and validation of the flood extent predicted by an urban flood inundation model. To date, research on such models has been hampered by lack of suitable distributed validation data. The study uses a 3m resolution TerraSAR-X image of a 1-in-150 year flood near Tewkesbury, UK, in 2007, for which contemporaneous aerial photography exists for validation. The DLR SETES SAR simulator was used in conjunction with airborne LiDAR data to estimate regions of the TerraSAR-X image in which water would not be visible due to radar shadow or layover caused by buildings and taller vegetation, and these regions were masked out in the flood detection process. A semi-automatic algorithm for the detection of floodwater was developed, based on a hybrid approach. Flooding in rural areas adjacent to the urban areas was detected using an active contour model (snake) region-growing algorithm seeded using the un-flooded river channel network, which was applied to the TerraSAR-X image fused with the LiDAR DTM to ensure the smooth variation of heights along the reach. A simpler region-growing approach was used in the urban areas, which was initialized using knowledge of the flood waterline in the rural areas. Seed pixels having low backscatter were identified in the urban areas using supervised classification based on training areas for water taken from the rural flood, and non-water taken from the higher urban areas. Seed pixels were required to have heights less than a spatially-varying height threshold determined from nearby rural waterline heights. Seed pixels were clustered into urban flood regions based on their close proximity, rather than requiring that all pixels in the region should have low backscatter. This approach was taken because it appeared that urban water backscatter values were corrupted in some pixels, perhaps due to contributions from side-lobes of strong reflectors nearby. The TerraSAR-X urban flood extent was validated using the flood extent visible in the aerial photos. It turned out that 76% of the urban water pixels visible to TerraSAR-X were correctly detected, with an associated false positive rate of 25%. If all urban water pixels were considered, including those in shadow and layover regions, these figures fell to 58% and 19% respectively. These findings indicate that TerraSAR-X is capable of providing useful data for the calibration and validation of urban flood inundation models.

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.

100 years on: a century of genetics

In 2006 we celebrated the centenary of a remarkable year that saw the birth of genetics as a scientific discipline. This birth had its origins in horticulture and was supervised by a remarkable Cambridge academic, accompanied by a loyal group of female colleagues who worked together in underfunded conditions with little institutional support. Despite this deprivation, they established the foundations of an ongoing revolution, with huge academic and commercial consequences that we can recognize today in the shape of genomics and its application to biomedicine.