Using decision trees to understand structure in missing data

Objectives Demonstrate the application of decision trees – classification and regression trees (CARTs), and their cousins, boosted regression trees (BRTs) – to understand structure in missing data. Setting Data taken from employees at three different industry sites in Australia. Participants 7915 observations were included. Materials and Methods The approach was evaluated using an occupational health dataset comprising results of questionnaires, medical tests, and environmental monitoring. Statistical methods included standard statistical tests and the ‘rpart’ and ‘gbm’ packages for CART and BRT analyses, respectively, from the statistical software ‘R’. A simulation study was conducted to explore the capability of decision tree models in describing data with missingness artificially introduced. Results CART and BRT models were effective in highlighting a missingness structure in the data, related to the Type of data (medical or environmental), the site in which it was collected, the number of visits and the presence of extreme values. The simulation study revealed that CART models were able to identify variables and values responsible for inducing missingness. There was greater variation in variable importance for unstructured compared to structured missingness. Discussion Both CART and BRT models were effective in describing structural missingness in data. CART models may be preferred over BRT models for exploratory analysis of missing data, and selecting variables important for predicting missingness. BRT models can show how values of other variables influence missingness, which may prove useful for researchers. Conclusion Researchers are encouraged to use CART and BRT models to explore and understand missing data.

Investigation of a suicide ideation risk profile in people with co-occurring depression and substance use disorder

Disengagement from services is common before suicide, hence identifying factors at treatment presentation that predict future suicidality is important. This article explores risk profiles for suicidal ideation among treatment seekers with depression and substance misuse. Participants completed assessments at baseline and 6 months. Baseline demographics, psychiatric history, and current symptoms were entered into a decision tree to predict suicidal ideation at follow-up. Sixty-three percent of participants at baseline and 43.5% at follow-up reported suicidal ideation. Baseline ideation most salient when psychiatric illness began before adulthood, increasing the rate of follow-up ideation by 16%. Among those without baseline ideation, dysfunctional attitudes were the most important risk factor, increasing rates of suicidal ideation by 35%. These findings provide evidence of factors beyond initial diagnoses that increase the likelihood of suicidal ideation and are worthy of clinical attention. In particular, providing suicide prevention resources to those with high dysfunctional attitudes may be beneficial.

Multi-Pattern Viterbi Algorithm for joint decoding of multiple speech patterns

Joint decoding of multiple speech patterns so as to improve speech recognition performance is important, especially in the presence of noise. In this paper, we propose a Multi-Pattern Viterbi algorithm (MPVA) to jointly decode and recognize multiple speech patterns for automatic speech recognition (ASR). The MPVA is a generalization of the Viterbi Algorithm to jointly decode multiple patterns given a Hidden Markov Model (HMM). Unlike the previously proposed two stage Constrained Multi-Pattern Viterbi Algorithm (CMPVA),the MPVA is a single stage algorithm. MPVA has the advantage that it cart be extended to connected word recognition (CWR) and continuous speech recognition (CSR) problems. MPVA is shown to provide better speech recognition performance than the earlier techniques: using only two repetitions of noisy speech patterns (-5 dB SNR, 10% burst noise), the word error rate using MPVA decreased by 28.5%, when compared to using individual decoding. (C) 2010 Elsevier B.V. All rights reserved.

Methods in general model localization

The aim of this study was to evaluate and test methods which could improve local estimates of a general model fitted to a large area. In the first three studies, the intention was to divide the study area into sub-areas that were as homogeneous as possible according to the residuals of the general model, and in the fourth study, the localization was based on the local neighbourhood. According to spatial autocorrelation (SA), points closer together in space are more likely to be similar than those that are farther apart. Local indicators of SA (LISAs) test the similarity of data clusters. A LISA was calculated for every observation in the dataset, and together with the spatial position and residual of the global model, the data were segmented using two different methods: classification and regression trees (CART) and the multiresolution segmentation algorithm (MS) of the eCognition software. The general model was then re-fitted (localized) to the formed sub-areas. In kriging, the SA is modelled with a variogram, and the spatial correlation is a function of the distance (and direction) between the observation and the point of calculation. A general trend is corrected with the residual information of the neighbourhood, whose size is controlled by the number of the nearest neighbours. Nearness is measured as Euclidian distance. With all methods, the root mean square errors (RMSEs) were lower, but with the methods that segmented the study area, the deviance in single localized RMSEs was wide. Therefore, an element capable of controlling the division or localization should be included in the segmentation-localization process. Kriging, on the other hand, provided stable estimates when the number of neighbours was sufficient (over 30), thus offering the best potential for further studies. Even CART could be combined with kriging or non-parametric methods, such as most similar neighbours (MSN).

Lahopuumäärän ennustaminen ja kartoitus lentokonelaserkeilauksen avulla

Lahopuun määrästä ja sijoittumisesta ollaan kiinnostuneita paitsi elinympäristöjen monimuotoisuuden, myös ilmakehän hiilen varastoinnin kannalta. Tutkimuksen tavoitteena oli kehittää aluepohjainen laserkeilausdataa hyödyntävä malli lahopuukohteiden paikantamiseksi ja lahopuun määrän estimoimiseksi. Samalla tutkittiin mallin selityskyvyn muuttumista mallinnettavan ruudun kokoa suurennettaessa. Tutkimusalue sijaitsi Itä-Suomessa Sonkajärvellä ja koostui pääasiassa nuorista hoidetuista talousmetsistä. Tutkimuksessa käytettiin harvapulssista laserkeilausdataa sekä kaistoittain mitattua maastodataa kuolleesta puuaineksesta. Aineisto jaettiin siten, että neljäsosa datasta oli käytössä mallinnusta varten ja loput varattiin valmiiden mallien testaamiseen. Lahopuun mallintamisessa käytettiin sekä parametrista että ei-parametrista mallinnusmenetelmää. Logistisen regression avulla erikokoisille (0,04, 0,20, 0,32, 0,52 ja 1,00 ha) ruuduille ennustettiin todennäköisyys lahopuun esiintymiselle. Muodostettujen mallien selittävät muuttujat valittiin 80 laserpiirteen ja näiden muunnoksien joukosta. Mallien selittävät muuttujat valittiin kolmessa vaiheessa. Aluksi muuttujia tarkasteltiin visuaalisesti kuvaamalla ne lahopuumäärän suhteen. Ensimmäisessä vaiheessa sopivimmiksi arvioitujen muuttujien selityskykyä testattiin mallinnuksen toisessa vaiheessa yhden muuttujan mallien avulla. Lopullisessa usean muuttujan mallissa selittävien muuttujien kriteerinä oli tilastollinen merkitsevyys 5 % riskitasolla. 0,20 hehtaarin ruutukoolle luotu malli parametrisoitiin muun kokoisille ruuduille. Logistisella regressiolla toteutetun parametrisen mallintamisen lisäksi, 0,04 ja 1,0 hehtaarin ruutukokojen aineistot luokiteltiin ei-parametrisen CART-mallinnuksen (Classification and Regression Trees) avulla. CARTmenetelmällä etsittiin aineistosta vaikeasti havaittavia epälineaarisia riippuvuuksia laserpiirteiden ja lahopuumäärän välillä. CART-luokittelu tehtiin sekä lahopuustoisuuden että lahopuutilavuuden suhteen. CART-luokituksella päästiin logistista regressiota parempiin tuloksiin ruutujen luokituksessa lahopuustoisuuden suhteen. Logistisella mallilla tehty luokitus parani ruutukoon suurentuessa 0,04 ha:sta(kappa 0,19) 0,32 ha:iin asti (kappa 0,38). 0,52 ha:n ruutukoolla luokituksen kappa-arvo kääntyi laskuun (kappa 0,32) ja laski edelleen hehtaarin ruutukokoon saakka (kappa 0,26). CART-luokitus parani ruutukoon kasvaessa. Luokitustulokset olivat logistista mallinnusta parempia sekä 0,04 ha:n (kappa 0,24) että 1,0 ha:n (kappa 0,52) ruutukoolla. CART-malleilla määritettyjen ruutukohtaisten lahopuutilavuuksien suhteellinen RMSE pieneni ruutukoon kasvaessa. 0,04 hehtaarin ruutukoolla koko aineiston lahopuumäärän suhteellinen RMSE oli 197,1 %, kun hehtaarin ruutukoolla vastaava luku oli 120,3 %. Tämän tutkimuksen tulosten perusteella voidaan todeta, että maastossa mitatun lahopuumäärän ja tutkimuksessa käytettyjen laserpiirteiden yhteys on pienellä ruutukoolla hyvin heikko, mutta vahvistuu hieman ruutukoon kasvaessa. Kun mallinnuksessa käytetty ruutukoko kasvaa, pienialaisten lahopuukeskittymien havaitseminen kuitenkin vaikeutuu. Tutkimuksessa kohteen lahopuustoisuus pystyttiin kartoittamaan kohtuullisesti suurella ruutukoolla, mutta pienialaisten kohteiden kartoittaminen ei onnistunut käytetyillä menetelmillä. Pienialaisten kohteiden paikantaminen laserkeilauksen avulla edellyttää jatkotutkimusta erityisesti tiheäpulssisen laserdatan käytöstä lahopuuinventoinneissa.

Combination antiretroviral therapy -associated lipodystrophy : insights into pathogenesis and treatment

Introduction: Combination antiretroviral therapy (cART) has decreased morbidity and mortality of individuals infected with human immunodeficiency virus type 1 (HIV-1). Its use, however, is associated with adverse effects which increase the patients risk of conditions such as diabetes and coronary heart disease. Perhaps the most stigmatizing side effect is lipodystrophy, i.e., the loss of subcutaneous adipose tissue (SAT) in the face, limbs and trunk while fat accumulates intra-abdominally and dorsocervically. The pathogenesis of cART-associated lipodystrophy is obscure. Nucleoside reverse transcriptase inhibitors (NRTI) have been implicated to cause lipoatrophy via mitochondrial toxicity. There is no known effective treatment for cART-associated lipodystrophy during unchanged antiretroviral regimen in humans, but in vitro data have shown uridine to abrogate NRTI-induced toxicity in adipocytes. Aims: To investigate whether i) cART or lipodystrophy associated with its use affect arterial stiffness; ii) lipoatrophic SAT is inflamed compared to non-lipoatrophic SAT; iii) abdominal SAT from patients with compared to those without cART-associated lipoatrophy differs with respect to mitochondrial DNA (mtDNA) content, adipose tissue inflammation and gene expression, and if NRTIs stavudine and zidovudine are associated with different degree of changes; iv) lipoatrophic abdominal SAT differs from preserved dorsocervical SAT with respect to mtDNA content, adipose tissue inflammation and gene expression in patients with cART-associated lipodystrophy and v) whether uridine can revert lipoatrophy and the associated metabolic disturbances in patients on stavudine or zidovudine based cART. Subjects and methods: 64 cART-treated patients with (n=45) and without lipodystrophy/-atrophy (n=19) were compared cross-sectionally. A marker of arterial stiffness, heart rate corrected augmentation index (AgIHR), was measured by pulse wave analysis. Body composition was measured by magnetic resonance imaging and dual-energy X-ray absorptiometry, and liver fat content by proton magnetic resonance spectroscopy. Gene expression and mtDNA content in SAT were assessed by real-time polymerase chain reaction and microarray. Adipose tissue composition and inflammation were assessed by histology and immunohistochemistry. Dorsocervical and abdominal SAT were studied. The efficacy and safety of uridine for the treatment of cART-associated lipoatrophy were evaluated in a randomized, double-blind, placebo-controlled 3-month trial in 20 lipoatrophic cART-treated patients. Results: Duration of antiretroviral treatment and cumulative exposure to NRTIs and protease inhibitors, but not the presence of cART-associated lipodystrophy, predicted AgIHR independent of age and blood pressure. Gene expression of inflammatory markers was increased in SAT of lipodystrophic as compared to non-lipodystrophic patients. Expression of genes involved in adipogenesis, triglyceride synthesis and glucose disposal was lower and of those involved in mitochondrial biogenesis, apoptosis and oxidative stress higher in SAT of patients with than without cART-associated lipoatrophy. Most changes were more pronounced in stavudine-treated than in zidovudine-treated individuals. Lipoatrophic SAT had lower mtDNA than SAT of non-lipoatrophic patients. Expression of inflammatory genes was lower in dorsocervical than in abdominal SAT. Neither depot had characteristics of brown adipose tissue. Despite being spared from lipoatrophy, dorsocervical SAT of lipodystrophic patients had lower mtDNA than the phenotypically similar corresponding depot of non-lipodystrophic patients. The greatest difference in gene expression between dorsocervical and abdominal SAT, irrespective of lipodystrophy status, was in expression of homeobox genes that regulate transcription and regionalization of organs during embryonal development. Uridine increased limb fat and its proportion of total fat, but had no effect on liver fat content and markers of insulin resistance. Conclusions: Long-term cART is associated with increased arterial stiffness and, thus, with higher cardiovascular risk. Lipoatrophic abdominal SAT is characterized by inflammation, apoptosis and mtDNA depletion. As mtDNA is depleted even in non-lipoatrophic dorsocervical SAT, lipoatrophy is unlikely to be caused directly by mtDNA depletion. Preserved dorsocervical SAT of patients with cART-associated lipodystrophy is less inflamed than their lipoatrophic abdominal SAT, and does not resemble brown adipose tissue. The greatest difference in gene expression between dorsocervical and abdominal SAT is in expression of transcriptional regulators, homeobox genes, which might explain the differential susceptibility of these adipose tissue depots to cART-induced toxicity. Uridine is able to increase peripheral SAT in lipoatrophic patients during unchanged cART.

Divergencies of Perception : The Possibilities of Merleau-Pontian Phenomenology in Analyses of Contemporary Art

Maurice Merleau-Ponty (1908-1961) has been known as the philosopher of painting. His interest in the theory of perception intertwined with the questions concerning the artist s perception, the experience of an artwork and the possible interpretations of the artwork. For him, aesthetics was not a sub-field of philosophy, and art was not simply a subject matter for the aesthetic experience, but a form of thinking. This study proposes an opening for a dialogue between Merleau-Pontian phenomenology and contemporary art. The thesis examines his phenomenology through certain works of contemporary art and presents readings of these artworks through his phenomenology. The thesis both shows the potentiality of a method, but also engages in the critical task of finding the possible limitations of his approach. The first part lays out the methodological and conceptual points of departure of Merleau-Ponty s phenomenological approach to perception as well as the features that determined his discussion on encountering art. Merleau-Ponty referred to the experience of perceiving art using the notion of seeing with (voir selon). He stressed a correlative reciprocity described in Eye and Mind (1961) as the switching of the roles of the visible and the painter. The choice of artworks is motivated by certain restrictions in the phenomenological readings of visual arts. The examined works include paintings by Tiina Mielonen, a photographic work by Christian Mayer, a film by Douglas Gordon and Philippe Parreno, and an installation by Monika Sosnowska. These works resonate with, and challenge, his phenomenological approach. The chapters with case studies take up different themes that are central to Merleau-Ponty s phenomenology: space, movement, time, and touch. All of the themes are interlinked with the examined artworks. There are also topics that reappear in the thesis, such as the notion of écart and the question of encountering the other. As Merleau-Ponty argued, the sphere of art has a particular capability to address our being in the world. The thesis presents an interpretation that emphasises the notion of écart, which refers to an experience of divergence or dispossession. The sudden dissociation, surprise or rupture that is needed in order for a meeting between the spectator and the artwork, or between two persons, to be possible. Further, the thesis suggests that through artworks it is possible to take into consideration the écart, the divergence, that defines our subjectivity.

Clan-based evolutionary approach for solving control problems

This paper presents a dan-based evolutionary approach for solving control problems. Three selected control problems, viz. linear-quadratic, harvest, and push-cart problems, are solved using the proposed approach. Results are compared with those of the evolutionary programming (EP) approach. In most of the cases, the proposed approach is successful in obtaining (near) optimal solutions for these selected problems.

An Optimal Fuzzy Logic Controller tuned with Artificial Immune System

In this paper, a method for the tuning the membership functions of a Mamdani type Fuzzy Logic Controller (FLC) using the Clonal Selection Algorithm(CSA) a model of the Artificial Immune System(AIS) paradigm is examined. FLC's are designed for two problems, firstly the linear cart centering problem and secondly the highly nonlinear inverted pendulum problem. The FLC tuned by AIS is compared with FLC tuned by GA. In order to check the robustness of the designed PLC's white noise was added to the system, further, the masses of the cart and the length and mass of the pendulum are changed. The PLC's were also tested in the presence of faulty rules. Finally, Kruskal Wallis test was performed to compare the performance of the GA and AIS. An insight into the algorithms are also given by studying the effect of the important parameters of GA and AIS.

Probabilistic inference for fast learning in control

A novel framework is provided for very fast model-based reinforcement learning in continuous state and action spaces. It requires probabilistic models that explicitly characterize their levels of condence. Within the framework, exible, non-parametric models are used to describe the world based on previously collected experience. It demonstrates learning on the cart-pole problem in a setting where very limited prior knowledge about the task has been provided. Learning progressed rapidly, and a good policy found after only a small number of iterations.

Survey of the fatty acid composition of Canadian beef: Backfat and longissimus lumborum muscle

[EN]A survey of Canadian retail beef was undertaken with emphasis on the trans fatty acid (TFA) and conjugated linoleic acid (CLA) isomers, and compared with current health recommendations. Thirty striploin steaks were collected in the winter and summer from major grocery stores in Calgary (Alberta, Canada). Steak fatty acid compositions (backfat and longissimus lumborum muscle analysed separately) showed minor seasonal differences with lower total saturates (PB0.05) and higher total monounsaturates (PB 0.01) in winter, but no differences in total polyunsaturated fatty acids. The ratio of n-6 and n-3 polyunsaturated fatty acid in longissimus lumborum averaged 5.8. The average TFA content in longissimus lumborum was 0.128 g 100 g_1 serving size, and 10t-18:1 was found to be the predominant isomer (32% of total trans), while vaccenic acid was second most abundant (15% of total trans). The CLA content in longissimus lumborum was similar to that of backfat, ranging from 0.43 to 0.60% of total fatty acids and rumenic acid represented 60% of total isomers. Overall, there is still room for improvement in the saturated, mono- and polyunsaturated fatty acid composition of Canadian beef to meet general dietary guidelines for human consumption and additional targets should include reducing 10t-18:1 while increasing both rumenic and vaccenic acids.

Using the system dynamics paradigm in teaching and learning technological university subjets

2nd International Conference on Education and New Learning Technologies

The Federal Fisheries Service, 1871–1940: Its Origins, Organization, and Accomplishments

The U.S. Fish Commission was initiated in 1871 with Spencer Fullerton Baird as the first U.S. Fish Commissioner as an independent entity. In 1903 it became a part of the new U.S. Department of Commerce and Labor and was renamed the Bureau of Fisheries, a name it retained when the Departments of Commerce and Labor were separated in 1912. The Bureau remained in the Commerce Department until 1941 when it was merged with the Biological Survey and placed in the Department of Interior as the U.S. Fish and Wildlife Service. It was a scientific agency with well conceived programs of action, and it provided knowledge, advice, and example to state governments and individuals with fisheries interests and needs. Its efforts were supported by timely international agreements which constituted the precedent for Federal interest in fishery matters. The Fisheries Service earned stature as an advisor through heavy emphasis on basic biological research. The lack of such knowledge was marked and universal in the 1870’s, but toward the end of that decade, strong steps had been taken to address those needs under Baird’s leadership. USFC research activities were conducted cooperatively with other prominent scientists in the United States and abroad. Biological stations were established, and the world’s first and most productive deepsea research vessel, the Albatross, was constructed, and its 40-year career gave a strong stimulus to the science of oceanography. Together, the agency’s scientists and facilities made important additions to the sum of human knowledge, derived principles of conservation which were the vital bases for effective regulatory legislation, conducted extensive fish cultural work, collected and disseminated fisheries statistics, and began important research in methods of fish harvesting, preservation, transportation, and marketing.

Condições de vida e mortalidade por doenças cardiovasculares na área de influência do Complexo Petroquímico do Estado do Rio de Janeiro - COMPERJ.

O objetivo desta tese foi identificar e caracterizar áreas com altas taxas de mortalidade por doenças do aparelho circulatório (DAC) e seus dois principais subgrupos de causas: as doenças isquêmicas do coração (DIC) e as doenças cerebrovasculares (DCV), entre os anos de 2008 e 2012 na área de influência do complexo petroquímico do estado do Rio de Janeiro COMPERJ, por meio de métodos estatísticos e sistemas de informações geográficas (SIG). Os resultados da investigação são apresentados na forma de dois manuscritos. O primeiro objetivou descrever o perfil da distribuição espacial da mortalidade por (DAC), caracterizar e predizer territórios com maior risco de morte por esta causa, com base em classificação das unidades espaciais por indicador de qualidade urbana (IQUmod). A análise multivariada foi realizada por meio do método conhecido como árvore de decisão e regressão, baseado em algoritmo CART para a obtenção do modelo preditivo para UVLs com diferentes riscos de mortalidade por DAC. O modelo desenvolvido foi capaz de discriminar sete conjuntos de UVLs, com diferentes taxas médias de mortalidades. O subconjunto que apresenta a maior taxa média (1037/100 mil hab.) apresenta 3 UVLs com mais de 75% de seus domicílios com abastecimento de água inadequado e valor de IQUmod acima de 0.6. Conclui-se que na área de influência do COMPERJ existem áreas onde a mortalidade por DAC se apresenta com maior magnitude e que a identificação dessas áreas pode auxiliar na elaboração, diagnóstico, prevenção e planejamento de ações de saúde direcionadas aos grupos mais susceptíveis. O segundo manuscrito teve por objetivo descrever o perfil da distribuição espacial da mortalidade por DIC e DCV em relação ao contexto socioambiental segundo áreas geográficas. O modelo de regressão linear de Poisson com parâmetro de estimação via quasi-verossimilhança foi usado para verificar associação entre as variáveis. Foram identificados como fatores de risco para mortalidade por DIC e DCV a variável relativa a melhor renda e maior distância entre domicílios e unidades de saúde; a proporção de domicílios em ruas pavimentadas aparece como fator de proteção. A distribuição espacial e as associações encontradas entre os desfechos e preditores sugerem que as populações residentes em localidades mais afastadas dos centros urbanos apresentam maiores taxas de mortalidade por DIC e DCV e que isto pode estar relacionado a contextos rurais de localização das residências e a distância geográfica destas populações aos serviços de saúde. Aponta-se para a necessidade de desenvolvimento de ações que propiciem maior amplitude no atendimento em saúde, no intuito da redução de eventos cardiovasculares mórbidos incidentes naquelas populações.

O lugar do sujeito nos Serviços Residenciais Terapêuticos

Esta pesquisa iniciou-se a partir do questionamento geral sobre o que faria a clínica psicanalítica no Serviço Residencial Terapêutico. Buscaram-se balizas na história da reforma psiquiátrica brasileira e em fundamentos da psicanálise aplicada a terapêutica a fim de discutir uma articulação entre essas duas áreas. É defendida a tese de que existe uma clínica necessária ao trabalho nos SRT, já que a mudança da realidade objetiva conseqüente à saída dos asilos não é suficiente para que os moradores sintam-se em casa. Além disso, é defendida a tese de que trata-se de psicanálise no seio de uma casa, mas não de um dispositivo específico ao SRT que possa ser definido a partir de critérios ou categorias da clínica, da terapêutica ou do tratamento. Questionou-se, então, como acontece a operação que faz com que se reconheça um espaço como íntimo ou o contrário, estranho. Ou seja, o sentimento de se estar em casa, a sensação de algo da ordem do familiar não é automática, sendo operação dependente de um trabalho subjetivo. Para Lacan, a casa do sujeito encontra-se em um ponto no Outro, situado por ele a partir de uma ausência, o φ. Na neurose esta cava é mobiliada, deixando furos, pela fantasia, ordenada pelo objeto a. Estas elaborações foram trabalhadas a partir de um caso. Foi discutido por fim o trabalho do cuidador e a supervisão nele implicada como o profissional que sustenta a brecha pela qual se pode abrir ao inconsciente para que o sujeito possa retornar a sua casa