Konputaezintasun frogapen batzuk diagonalizazio teknika erabiliz

Konputagarritasunaren Teoriaren oinarriak lehenengo ordenadoreak azaldu aurretik (40. hamarkadaren bukaera aldera) ezarri ziren, eta ziztu biziko eta etenik gabeko eraldaketek aldatzea lortu ez duten oinarriak dira. Alan Mathison Turing-ek jadanik garai hartan frogatu zuen, ahalik eta potentzia handienekoa imajinatuta ere, inolako ordenadorek ebatzi ezingo zituen zenbait gai edo arazo bazeudela. Balizko algoritmorik ez duten problema horiek, konputaezinak deitzen ditugunak, ez dira salbuespenak eta adibide ugari aurki dezakegu. Programen portaeraren inguruan planteatzen diren problemen artean, asko konputaezinak dira. Familia horretako kide ezagunena, zalantzarik gabe, geratze problema da: sarrerako datu zehatz batzuk hartzerakoan, programa bat begizta infinituan geratuko ote den era orokorrean erabakitzeko algoritmorik ez dago. Problema baten konputaezintasuna frogatzeko, hau ebatziko duen algoritmo zehatz bat existitzen ez dela ziurtatuko duen argumentu logikoa behar dugu, edo beste era batera esanda, existitzen diren algoritmoak problema hori ebazteko gai izango ez direla egiaztatuko duen argumentua. Izaera unibertsaleko argumentu hori ezartzea ez da batere erraza izaten, eta normalean, absurduraino eramandako frogapen batekin erlazionatuta egon ohi da. Helburu hori lortzeko zenbait teknika daude. Diagonalizazioaren teknika horien artean oinarrizkoena da, eta nahiko ezaguna, ez baita Informatika Teorikoaren tresna espezifikoa. Dokumentu honen helburua ez da teknika bera azaldu edo deskribatzea, ezaguntzat hartzen baita, zailtasun maila desberdineko hainbat adibideren bitartez argitzea baizik.

Identification of nonlinear systems using generalized kernel models

Nonlinear system identification is considered using a generalized kernel regression model. Unlike the standard kernel model, which employs a fixed common variance for all the kernel regressors, each kernel regressor in the generalized kernel model has an individually tuned diagonal covariance matrix that is determined by maximizing the correlation between the training data and the regressor using a repeated guided random search based on boosting optimization. An efficient construction algorithm based on orthogonal forward regression with leave-one-out (LOO) test statistic and local regularization (LR) is then used to select a parsimonious generalized kernel regression model from the resulting full regression matrix. The proposed modeling algorithm is fully automatic and the user is not required to specify any criterion to terminate the construction procedure. Experimental results involving two real data sets demonstrate the effectiveness of the proposed nonlinear system identification approach.

Humoral immunity following chickenpox is influenced by geography and ethnicity

Objectives To investigate the contribution of ethnicity and geographical location to varicella-zoster virus (VZV) serostatus and antibody concentrations. Methods The presence and concentrations of antibodies to VZV were measured in 639 Bangladeshi women born in Bangladesh (BBB), 94 Bangladeshi women born in the UK (BUK) and 262 White women born in the UK (WUK). The results were anaylsed in relation to demographic and social data. Results BBB women were significantly less likely to be VZV seropositive at all ages than both BUK and WUK women. However, the odds of a Bangladeshi-born woman being seropositive increased by 1.04 for each year under the age of 15 spent in the UK. In contrast, antibody concentrations were significantly lower in ethnic Bangladeshi women, irrespective of country of birth. White, but not Bangladeshi women, showed evidence of antibody boosting over time despite the latter having more exposure to children. Conclusion Geographical location during childhood is the major influence on age of primary infection with VZV while the level of antibody is related to ethnicity. Since the risk of re-infection with VZV following both natural infection and vaccination is increased as antibody concentrations fall, these results have implications for VZV vaccination programmes particularly in non-White populations.

Researcher development: the impact of undergraduate research opportunity programmes on students in the UK

Undergraduate research opportunity programmes (UROP) are common in North America where research has confirmed their benefits. These schemes are gaining ground in the UK, and this article provides evidence for how UK students are benefiting from the experience. Results suggest UROP makes a significant contribution to the research capabilities and confidence of participating students, boosting their understanding of both research and their own subjects. Whilst offering considerable benefits to student learning, there is no evidence that UROP schemes on their current small scale attract additional students to postgraduate research, since the majority that participate are already interested in postgraduate study. However, at an individual level, most students report increased confidence and appreciation of the realities of the research process, and desire to progress on to postgraduate study following the placement, indicating that schemes may have the potential to cultivate new research confidence and interest if expanded.

Farm productivity and efficiency in rural Bangladesh: the role of education revisited

This article reassesses the debate over the role of education in farm production in Bangladesh using a large dataset on rice producing households from 141 villages. Average and stochastic production frontier functions are estimated to ascertain the effect of education on productivity and efficiency. A full set of proxies for farm education stock variables are incorporated to investigate the ‘internal’ as well as ‘external’ returns to education. The external effect is investigated in the context of rural neighbourhoods. Our analysis reveals that in addition to raising rice productivity and boosting potential output, household education significantly reduces production inefficiencies. However, we are unable to find any evidence of the externality benefit of schooling – neighbour's education does not matter in farm production. We discuss the implication of these findings for rural education programmes in Bangladesh.

The identity of crop pollinators helps target conservation for improved ecosystem services

Insect pollinated mass flowering crops are becoming more widespread and there is a need to understand which insects are primarily responsible for the pollination of these crops so conservation measures can be appropriately targeted in the face of pollinator declines. This study used field surveys in conjunction with cage manipulations to identify the relative contributions of different pollinator taxa to the pollination of two widespread flowering crops, field beans and oilseed rape. Flower visiting pollinator communities observed in the field were distinct for each crop; while field beans were visited primarily by a few bumblebee species, multiple pollinator taxa visited oilseed, and the composition of this pollinator community was highly variable spatially and temporally. Neither pollinator community, however, appears to be meeting the demands of crops in our study regions. Cage manipulations showed that multiple taxa can effectively pollinate both oilseed and field beans, but bumblebees are particularly effective bean pollinators. Combining field observations and cage manipulations demonstrated that the pollination demands of these two mass flowering crops are highly contrasting, one would benefit from management to increase the abundance of some key taxa, whilst for the other, boosting overall pollinator abundance and diversity would be more appropriate. Our findings highlight the need for crop specific mitigation strategies that are targeted at conserving specific pollinator taxa (or group of taxa) that are both active and capable of crop pollination in order to reduce pollination deficits and meet the demands of future crop production.

Climate-driven spatial mismatches between British orchards and their pollinators: increased risks of pollination deficits

Understanding how climate change can affect crop-pollinator systems helps predict potential geographical mismatches between a crop and its pollinators, and therefore identify areas vulnerable to loss of pollination services. We examined the distribution of orchard species (apples, pears, plums and other top fruits) and their pollinators in Great Britain, for present and future climatic conditions projected for 2050 under the SRES A1B Emissions Scenario. We used a relative index of pollinator availability as a proxy for pollination service. At present there is a large spatial overlap between orchards and their pollinators, but predictions for 2050 revealed that the most suitable areas for orchards corresponded to low pollinator availability. However, we found that pollinator availability may persist in areas currently used for fruit production, but which are predicted to provide sub-optimal environmental suitability for orchard species in the future. Our results may be used to identify mitigation options to safeguard orchard production against the risk of pollination failure in Great Britain over the next 50 years; for instance choosing fruit tree varieties that are adapted to future climatic conditions, or boosting wild pollinators through improving landscape resources. Our approach can be readily applied to other regions and crop systems, and expanded to include different climatic scenarios.

Global networks, cities and economic performance: Observations from an analysis of cities in Europe and the USA

The network paradigm has been highly influential in spatial analysis in the globalisation era. As economies across the world have become increasingly integrated, so-called global cities have come to play a growing role as central nodes in the networked global economy. The idea that a city’s position in global networks benefits its economic performance has resulted in a competitive policy focus on promoting the economic growth of cities by improving their network connectivity. However, in spite of the attention being given to boosting city connectivity little is known about whether this directly translates to improved city economic performance and, if so, how well connected a city needs to be in order to benefit from this. In this paper we test the relationship between network connectivity and economic performance between 2000 and 2008 for cities with over 500,000 inhabitants in Europe and the USA to inform European policy.

ENSO and greenhouse warming

The El Niño/Southern Oscillation (ENSO) is the dominant climate phenomenon affecting extreme weather conditions worldwide. Its response to greenhouse warming has challenged scientists for decades, despite model agreement on projected changes in mean state. Recent studies have provided new insights into the elusive links between changes in ENSO and in the mean state of the Pacific climate. The projected slow-down in Walker circulation is expected to weaken equatorial Pacific Ocean currents, boosting the occurrences of eastward-propagating warm surface anomalies that characterize observed extreme El Niño events. Accelerated equatorial Pacific warming, particularly in the east, is expected to induce extreme rainfall in the eastern equatorial Pacific and extreme equatorward swings of the Pacific convergence zones, both of which are features of extreme El Niño. The frequency of extreme La Niña is also expected to increase in response to more extreme El Niños, an accelerated maritime continent warming and surface-intensified ocean warming. ENSO-related catastrophic weather events are thus likely to occur more frequently with unabated greenhouse-gas emissions. But model biases and recent observed strengthening of the Walker circulation highlight the need for further testing as new models, observations and insights become available.

Immune Responses and Therapeutic Antitumor Effects of an Experimental DNA Vaccine Encoding Human Papillomavirus Type 16 Oncoproteins Genetically Fused to Herpesvirus Glycoprotein D

Recombinant adenovirus or DNA vaccines encoding herpes simplex virus type 1 (HSV-1) glycoprotein D (gD) genetically fused to human papillomavirus type 16 (HPV-16) oncoproteins (E5, E6, and E7) induce antigen-specific CD8(+) T-cell responses and confer preventive resistance to transplantable murine tumor cells (TC-1 cells). In the present report, we characterized some previously uncovered aspects concerning the induction of CD8(+) T-cell responses and the therapeutic anticancer effects achieved in C57BL/6 mice immunized with pgD-E7E6E5 previously challenged with TC-1 cells. Concerning the characterization of the immune responses elicited in mice vaccinated with pgD-E7E6E5, we determined the effect of the CD4(+) T-cell requirement, longevity, and dose-dependent activation on the E7-specific CD8(+) T-cell responses. In addition, we determined the priming/boosting properties of pgD-E7E6E5 when used in combination with a recombinant serotype 68 adenovirus (AdC68) vector encoding the same chimeric antigen. Mice challenged with TC-1 cells and then immunized with three doses of pgD-E7E6E5 elicited CD8(+) T-cell responses, measured by intracellular gamma interferon (IFN-gamma) and CD107a accumulation, to the three HPV-16 oncoproteins and displayed in vivo antigen-specific cytolytic activity, as demonstrated with carboxyfluorescein diacetate succinimidyl ester (CFSE)-labeled target cells pulsed with oligopeptides corresponding to the H-2D(b)-restricted immunodominant epitopes of the E7, E6, or E5 oncoprotein. Up to 70% of the mice challenged with 5 x 10(5) TC-1 cells and immunized with pgD-E7E6E5 controlled tumor development even after 3 days of tumor cell challenge. In addition, coadministration of pgD-E7E6E5 with DNA vectors encoding pGM-CSF or interleukin-12 (IL-12) enhanced the therapeutic antitumor effects for all mice challenged with TC-1 cells. In conclusion, the present results expand our previous knowledge on the immune modulation properties of the pgD-E7E6E5 vector and demonstrate, for the first time, the strong antitumor effects of the DNA vaccine, raising promising perspectives regarding the development of immunotherapeutic reagents for the control of HPV-16-associated tumors.

Naturally acquired inhibitory antibodies to Plasmodium vivax Duffy binding protein are short-lived and allele-specific following a single malaria infection

The Duffy binding protein of Plasmodium vivax (DBP) is a critical adhesion ligand that participates in merozoite invasion of human Duffy-positive erythrocytes. A small outbreak of P. vivax malaria, in a village located in a non-malarious area of Brazil, offered us an opportunity to investigate the DBP immune responses among individuals who had their first and brief exposure to malaria. Thirty-three individuals participated in the five cross-sectional surveys, 15 with confirmed P. vivax infection while residing in the outbreak area (cases) and 18 who had not experienced malaria (non-cases). In the present study, we found that only 20% (three of 15) of the individuals who experienced their first P. vivax infection developed an antibody response to DBP; a secondary boosting can be achieved with a recurrent P. vivax infection. DNA sequences from primary/recurrent P. vivax samples identified a single dbp allele among the samples from the outbreak area. To investigate inhibitory antibodies to the ligand domain of the DBP (cysteine-rich region II, DBP(II)), we performed in vitro assays with mammalian cells expressing DBP(II) sequences which were homologous or not to those from the outbreak isolate. In non-immune individuals, the results of a 12-month follow-up period provided evidence that naturally acquired inhibitory antibodies to DBP(II) are short-lived and biased towards a specific allele.

Machine Learning Algorithms in Heavy Process Manufacturing

In a global economy, manufacturers mainly compete with cost efficiency of production, as the price of raw materials are similar worldwide. Heavy industry has two big issues to deal with. On the one hand there is lots of data which needs to be analyzed in an effective manner, and on the other hand making big improvements via investments in cooperate structure or new machinery is neither economically nor physically viable. Machine learning offers a promising way for manufacturers to address both these problems as they are in an excellent position to employ learning techniques with their massive resource of historical production data. However, choosing modelling a strategy in this setting is far from trivial and this is the objective of this article. The article investigates characteristics of the most popular classifiers used in industry today. Support Vector Machines, Multilayer Perceptron, Decision Trees, Random Forests, and the meta-algorithms Bagging and Boosting are mainly investigated in this work. Lessons from real-world implementations of these learners are also provided together with future directions when different learners are expected to perform well. The importance of feature selection and relevant selection methods in an industrial setting are further investigated. Performance metrics have also been discussed for the sake of completion.

Proposta de uma arquitetura especial para simulação lógica

O objetivo deste trabalho é a proposta de uma arquitetura especial para simulação lógica (AESL). As técnicas e modelos utilizados no processo de simulação lógica são brevemente revistos. É definida uma taxonomia para AESL sob a qual são analisadas diversas propostas de AESL relatadas na literatura. Uma taxonomia já existente é comparada com a proposta. A AESL definida é programável para diferentes algoritmos de simulação lógica. O detalhamento da AESL é, então, incrementado pela implementação de um algoritmo particular. Uma linguagem de simulação discreta é utilizada na construção de um modelo da arquitetura. Os resultados da simulação deste modelo permitem avaliar o desempenho da AESL e otimizar sua estrutura. Uma comparação com outras arquiteturas conclui a análise.

Dimensionamento de pilares esbeltos de concreto armado

O objetivo deste trabalho é o dimensionamento de pilares esbeltos de concreto armado, sob cargas de curta e longa duração, baseado numa análise realística das deformações do mesmo. Apresenta-se três algoritmos numéricos para a obtencão das relações momento fletor-esforço normal-curvatura de uma seção arbitrária de concreto armado, sob flexo-compressão normal. Inclue-se as deformações específicas de fluência e retração do concreto na análise, através de uma alteração nas referidas relações. Apresenta-se alguns critérios de normas, relativos ao dimensionamento de pilares esbeltos de concreto armado e uma comparação dos mesmos, entre si e com o algoritmo numérico desenvolvido. Considerações da NB-1/78 relativas ao projeto de pilares são analisadas, verificando o nivel da precisão obtida. Um procedimento simplificado para a inclusão da fluência do concreto no dimensionamento, proposto pelo CEB, é testado e uma solução para pilares de concreto armado com engastamento elástico simétrico é apresentada, para verificar o nível: do erro cometido ao se estender o conceito de comprimento de flambagem a pilares de concreto armado. Uma série de exemplos experimentais são apresentados, onde a solução numérica para o dimensionamento tem sua precisão verificada. Diversas tabelas foram desenvolvidas para o dimensionamento de pilares esbeltos com secão transversal retangular e armadura simétrica. Todo o estudo é restrito ao caso de flexo-compressão normal.

Paralelização de métodos de resolução de sistemas lineares esparsos com o DECK em um Cluster de PCs

O objetivo desta dissertação é a paralelização e a avaliação do desempenho de alguns métodos de resolução de sistemas lineares esparsos. O DECK foi utilizado para implementação dos métodos em um cluster de PCs. A presente pesquisa é motivada pela vasta utilização de Sistemas de Equações Lineares em várias áreas científicas, especialmente, na modelagem de fenômenos físicos através de Equações Diferenciais Parciais (EDPs). Nessa área, têm sido desenvolvidas pesquisas pelo GMC-PAD – Grupo de Matemática da Computação e Processamento de Alto Desempenho da UFRGS, para as quais esse trabalho vem contribuindo. Outro fator de motivação para a realização dessa pesquisa é a disponibilidade de um cluster de PCs no Instituto de Informática e do ambiente de programação paralela DECK – Distributed Execution and Communication Kernel. O DECK possibilita a programação em ambientes paralelos com memória distribuída e/ou compartilhada. Ele está sendo desenvolvido pelo grupo de pesquisas GPPD – Grupo de Processamento Paralelo e Distribuído e com a paralelização dos métodos, nesse ambiente, objetiva-se também validar seu funcionamento e avaliar seu potencial e seu desempenho. Os sistemas lineares originados pela discretização de EDPs têm, em geral, como características a esparsidade e a numerosa quantidade de incógnitas. Devido ao porte dos sistemas, para a resolução é necessária grande quantidade de memória e velocidade de processamento, característicos de computações de alto desempenho. Dois métodos de resolução foram estudados e paralelizados, um da classe dos métodos diretos, o Algoritmo de Thomas e outro da classe dos iterativos, o Gradiente Conjugado. A forma de paralelizar um método é completamente diferente do outro. Isso porque o método iterativo é formado por operações básicas de álgebra linear, e o método direto é formado por operações elementares entre linhas e colunas da matriz dos coeficientes do sistema linear. Isso permitiu a investigação e experimentação de formas distintas de paralelismo. Do método do Gradiente Conjugado, foram feitas a versão sem précondicionamento e versões pré-condicionadas com o pré-condicionador Diagonal e com o pré-condicionador Polinomial. Do Algoritmo de Thomas, devido a sua formulação, somente a versão básica foi feita. Após a paralelização dos métodos de resolução, avaliou-se o desempenho dos algoritmos paralelos no cluster, através da realização de medidas do tempo de execução e foram calculados o speedup e a eficiência. As medidas empíricas foram realizadas com variações na ordem dos sistemas resolvidos e no número de nodos utilizados do cluster. Essa avaliação também envolveu a comparação entre as complexidades dos algoritmos seqüenciais e a complexidade dos algoritmos paralelos dos métodos. Esta pesquisa demonstra o desempenho de métodos de resolução de sistemas lineares esparsos em um ambiente de alto desempenho, bem como as potencialidades do DECK. Aplicações que envolvam a resolução desses sistemas podem ser realizadas no cluster, a partir do que já foi desenvolvido, bem como, a investigação de précondicionadores, comparação do desempenho com outros métodos de resolução e paralelização dos métodos com outras ferramentas possibilitando uma melhor avaliação do DECK.