Orthogonal forward selection for constructing the radial basis function network with tunable nodes

An orthogonal forward selection (OFS) algorithm based on the leave-one-out (LOO) criterion is proposed for the construction of radial basis function (RBF) networks with tunable nodes. This OFS-LOO algorithm is computationally efficient and is capable of identifying parsimonious RBF networks that generalise well. Moreover, the proposed algorithm is fully automatic and the user does not need to specify a termination criterion for the construction process.

Kernel classifier construction using orthogonal forward selection and boosting with Fisher ratio class separability measure

A greedy technique is proposed to construct parsimonious kernel classifiers using the orthogonal forward selection method and boosting based on Fisher ratio for class separability measure. Unlike most kernel classification methods, which restrict kernel means to the training input data and use a fixed common variance for all the kernel terms, the proposed technique can tune both the mean vector and diagonal covariance matrix of individual kernel by incrementally maximizing Fisher ratio for class separability measure. An efficient weighted optimization method is developed based on boosting to append kernels one by one in an orthogonal forward selection procedure. Experimental results obtained using this construction technique demonstrate that it offers a viable alternative to the existing state-of-the-art kernel modeling methods for constructing sparse Gaussian radial basis function network classifiers. that generalize well.

Modification of enrofloxacin treatment regimens for poultry experimentally infected with Salmonella enterica serovar Typhimurium DT104 to minimize selection of resistance

We hypothesized that higher doses of fluoroquinolones for a shorter duration could maintain efficacy (as measured by reduction in bacterial count) while reducing selection in chickens of bacteria with reduced susceptibility. Chicks were infected with Salmonella enterica serovar Typhimurium DT104 and treated 1 week later with enrofloxacin at the recommended dose for 5 days (water dose adjusted to give 10 mg/kg of body weight of birds or equivalence, i.e., water at 50 ppm) or at 2.5 or 5 times the recommended dose for 2 days or 1 day, respectively. The dose was delivered continuously (ppm) or pulsed in the water (mg/kg) or by gavage (mg/kg). In vitro in sera, increasing concentrations of 0.5 to 8 mu g/ml enrofloxacin correlated with increased activity. In vivo, the efficacy of the 1-day treatment was significantly less than that of the 2- and 5-day treatments. The 2-day treatments showed efficacy similar to that of the 5-day treatment in all but one repeat treatment group and significantly (P < 0.01) reduced the Salmonella counts. Dosing at 2.5x the recommended dose and pulsed dosing both increased the peak antibiotic concentrations in cecal contents, liver, lung, and sera as determined by high-pressure liquid chromatography. There was limited evidence that shorter treatment regimens (in particular the 1-day regimen) selected for fewer strains with reduced susceptibility. In conclusion, the 2-day treatment would overall require a shorter withholding time than the 5-day treatment and, in view of the increased peak antibiotic concentrations, may give rise to improved efficacy, in particular for treating respiratory and systemic infections. However, it would be necessary to validate the 2-day regimen in a field situation and in particular against respiratory and systemic infections to validate or refute this hypothesis.

Impact of HPA nonlinearity on MIMO systems with quantized equal gain transmission

In this paper, we investigate the performance of multiple-input multiple-output (MIMO) transmit beamforming (TB) systems in the presence of nonlinear high-power amplifiers (HPAs). Due to the suboptimality of maximal ratio transmission/maximal ratio combining (MRT/MRC) under HPA nonlinearity, quantized equal gain transmission (QEGT) is suggested as a feasible TB scheme. The effect of HPA nonlinearity on the performance of MIMO QEGT/MRC is evaluated in terms of the average symbol error probability (SEP) and system capacity, considering transmission over uncorrelated quasi-static frequency-flat Rayleigh fading channels. Numerical results are provided and show the effects of several system parameters, such as the parameters of nonlinear HPA, cardinality of the beamforming weight vector codebook, and modulation order of quadrature amplitude modulation (QAM), on performance.

Olfactory selection of Plantago lanceolata by snails declines with seedling age

Background and Aims Despite recent recognition that (1) plant–herbivore interactions during the establishment phase, (2) ontogenetic shifts in resource allocation and (3) herbivore response to plant volatile release are each pivotal to a comprehensive understanding of plant defence, no study has examined how herbivore olfactory response varies during seedling ontogeny. Methods Using a Y-tube olfactometer we examined snail (Helix aspersa) olfactory response to pellets derived from macerated Plantago lanceolata plants harvested at 1, 2, 3, 4, 5, 6 and 8 weeks of age to test the hypothesis that olfactory selection of plants by a generalist herbivore varies with plant age. Plant volatiles were collected for 10 min using solid-phase microextraction technique on 1- and 8-week-old P. lanceolata pellets and analysed by gas chromatography coupled with a mass spectrometer. Key Results Selection of P. lanceolata was strongly negatively correlated with increasing age; pellets derived from 1-week-old seedlings were three times more likely to be selected as those from 8-week-old plants. Comparison of plant selection experiments with plant volatile profiles from GC/MS suggests that patterns of olfactory selection may be linked to ontogenetic shifts in concentrations of green leaf volatiles and ethanol (and its hydrolysis derivatives). Conclusions Although confirmatory of predictions made by contemporary plant defence theory, this is the first study to elucidate a link between seedling age and olfactory selection by herbivores. As a consequence, this study provides a new perspective on the ontogenetic expression of seedling defence, and the role of seedling herbivores, particularly terrestrial molluscs, as selective agents in temperate plant communities.

Microhabitat Selection of three Forest Understory Birds in the Brazilian Atlantic Rainforest

When assessing fragmentation effects on species, not only habitat preferences on the landscape scale, but also microhabitat selection is an important factor to consider, as microhabitat is also affected by habitat disturbance, but nevertheless essential for species for foraging, nesting and sheltering. In the Atlantic Rainforest of Brazil we examined microhabitat selection of six Pyriglena leucoptera (white-shouldered fire-eye), 10 Sclerurus scansor (rufous-breasted leaftosser), and 30 Chiroxiphia caudata (blue manakin). We radio-tracked the individuals between May 2004 and February 2005 to gain home ranges based on individual fixed kernels. Vegetation structures in core plots and fringe plots were compared. In C. caudata, we additionally assessed the influence of behavioural traits on microhabitat selection. Further, we compared microhabitat structures in the fragmented forest with those in the contiguous, and contrasted the results with the birds` preferences. Pyriglena leucoptera preferred liana tangles that were more common in the fragmented forest, whereas S. scansor preferred woody debris, open forest floor (up to 0.5 m), and a thin closed leaf litter cover which all occurred significantly more often in the contiguous forest. Significant differences were detected in C. caudata for vegetation densities in the different strata; the distance of core plots to the nearest lek site was significantly influenced by sex and age. However, core sites of C. caudata in fragmented and contiguous forests showed no significant differences in structure. Exploring microhabitat selection and behavior may greatly support the understanding of habitat selection of species and their susceptibility to fragmentation on the landscape scale.

Towards improving cluster-based feature selection with a simplified silhouette filter

This paper proposes a filter-based algorithm for feature selection. The filter is based on the partitioning of the set of features into clusters. The number of clusters, and consequently the cardinality of the subset of selected features, is automatically estimated from data. The computational complexity of the proposed algorithm is also investigated. A variant of this filter that considers feature-class correlations is also proposed for classification problems. Empirical results involving ten datasets illustrate the performance of the developed algorithm, which in general has obtained competitive results in terms of classification accuracy when compared to state of the art algorithms that find clusters of features. We show that, if computational efficiency is an important issue, then the proposed filter May be preferred over their counterparts, thus becoming eligible to join a pool of feature selection algorithms to be used in practice. As an additional contribution of this work, a theoretical framework is used to formally analyze some properties of feature selection methods that rely on finding clusters of features. (C) 2011 Elsevier Inc. All rights reserved.

Model selection, estimation and forecasting in VAR models with short-run and long-run restrictions

We study the joint determination of the lag length, the dimension of the cointegrating space and the rank of the matrix of short-run parameters of a vector autoregressive (VAR) model using model selection criteria. We consider model selection criteria which have data-dependent penalties for a lack of parsimony, as well as the traditional ones. We suggest a new procedure which is a hybrid of traditional criteria and criteria with data-dependant penalties. In order to compute the fit of each model, we propose an iterative procedure to compute the maximum likelihood estimates of parameters of a VAR model with short-run and long-run restrictions. Our Monte Carlo simulations measure the improvements in forecasting accuracy that can arise from the joint determination of lag-length and rank, relative to the commonly used procedure of selecting the lag-length only and then testing for cointegration.

Model selection, estimation and forecasting in VAR models with short-run and long-run restrictions

We study the joint determination of the lag length, the dimension of the cointegrating space and the rank of the matrix of short-run parameters of a vector autoregressive (VAR) model using model selection criteria. We consider model selection criteria which have data-dependent penalties as well as the traditional ones. We suggest a new two-step model selection procedure which is a hybrid of traditional criteria and criteria with data-dependant penalties and we prove its consistency. Our Monte Carlo simulations measure the improvements in forecasting accuracy that can arise from the joint determination of lag-length and rank using our proposed procedure, relative to an unrestricted VAR or a cointegrated VAR estimated by the commonly used procedure of selecting the lag-length only and then testing for cointegration. Two empirical applications forecasting Brazilian inflation and U.S. macroeconomic aggregates growth rates respectively show the usefulness of the model-selection strategy proposed here. The gains in different measures of forecasting accuracy are substantial, especially for short horizons.

Model selection, estimation and forecasting in VAR models with short-run and long-run restrictions

We study the joint determination of the lag length, the dimension of the cointegrating space and the rank of the matrix of short-run parameters of a vector autoregressive (VAR) model using model selection criteria. We consider model selection criteria which have data-dependent penalties as well as the traditional ones. We suggest a new two-step model selection procedure which is a hybrid of traditional criteria and criteria with data-dependant penalties and we prove its consistency. Our Monte Carlo simulations measure the improvements in forecasting accuracy that can arise from the joint determination of lag-length and rank using our proposed procedure, relative to an unrestricted VAR or a cointegrated VAR estimated by the commonly used procedure of selecting the lag-length only and then testing for cointegration. Two empirical applications forecasting Brazilian in ation and U.S. macroeconomic aggregates growth rates respectively show the usefulness of the model-selection strategy proposed here. The gains in di¤erent measures of forecasting accuracy are substantial, especially for short horizons.

Model selection, estimation and forecasting in VAR models with short-run and long-run restrictions

We study the joint determination of the lag length, the dimension of the cointegrating space and the rank of the matrix of short-run parameters of a vector autoregressive (VAR) model using model selection criteria. We suggest a new two-step model selection procedure which is a hybrid of traditional criteria and criteria with data-dependant penalties and we prove its consistency. A Monte Carlo study explores the finite sample performance of this procedure and evaluates the forecasting accuracy of models selected by this procedure. Two empirical applications confirm the usefulness of the model selection procedure proposed here for forecasting.

Signaling in dynamic markets with adverse selection

Nesta dissertação, consideram-se trocas em mercados descentralizados com seleção adversa. Diferentemente da literatura até o momento, supomos que vendedores informados (e não compradores desinformados) fazem ofertas take-it-or-leave-it, de forma que sinalização através de preços é possível. Estabelecemos uma caracterização parcial do conjunto de equilíbrio, encontramos condições necessárias e suficientes para a existência de um equilíbrio e mostramos que todo equilíbrio apresenta sinalização se o problema de seleção adversa for suficientemente severo. Além disso, provamos o resultado surpreendente que o maior bem-estar atingido em equilíbrio é invariante às fricções do mercado. Também apresentamos condições necessárias e suficientes para a existência de equilíbrios separantes, que caracterizamos completamente. Mostramos que o conjunto de payoffs associados a equilíbrios separantes é invariante às fricções. Concluímos com uma caracterização completa do conjunto de equilíbrio com apenas dois tipos, e comparamos nossos resultados com os de Moreno e Wooders (2010), que analisam o caso em que compradores têm todo o poder de mercado. Nossos resultados mostram que sinalização através dos preços tem um impacto não trivial tanto nos resultados do mercado quanto no bem-estar.

Adverse selection with endogenous entry

Usando como base o ambiente descrito em Moreno e Wooders (2010), neste trabalho, analisamos trocas em um ambiente dinâmico, descentralizado e com seleção adversa. Ao contrário dos autores e da literatura, não consideramos a proporção de ativos de alta qualidade entrantes como independente das características do mercado. Desse modo, adaptamos o modelo dinâmico básico de seleção adversa para incorporar a decisão do vendedor sobre a possibilidade de pagar ou não um preço e transformar seu ativo de baixa qualidade em um ativo de alta qualidade antes de entrar no mercado. E, sob essas condições, mostramos que o bem-estar pode se comportar de maneira diferente do modelo tradicional.