Multi-scale support vector algorithms for hot spot detection and modelling

The algorithmic approach to data modelling has developed rapidly these last years, in particular methods based on data mining and machine learning have been used in a growing number of applications. These methods follow a data-driven methodology, aiming at providing the best possible generalization and predictive abilities instead of concentrating on the properties of the data model. One of the most successful groups of such methods is known as Support Vector algorithms. Following the fruitful developments in applying Support Vector algorithms to spatial data, this paper introduces a new extension of the traditional support vector regression (SVR) algorithm. This extension allows for the simultaneous modelling of environmental data at several spatial scales. The joint influence of environmental processes presenting different patterns at different scales is here learned automatically from data, providing the optimum mixture of short and large-scale models. The method is adaptive to the spatial scale of the data. With this advantage, it can provide efficient means to model local anomalies that may typically arise in situations at an early phase of an environmental emergency. However, the proposed approach still requires some prior knowledge on the possible existence of such short-scale patterns. This is a possible limitation of the method for its implementation in early warning systems. The purpose of this paper is to present the multi-scale SVR model and to illustrate its use with an application to the mapping of Cs137 activity given the measurements taken in the region of Briansk following the Chernobyl accident.

A Data mining approach to indirect inference

Consider a model with parameter phi, and an auxiliary model with parameter theta. Let phi be a randomly sampled from a given density over the known parameter space. Monte Carlo methods can be used to draw simulated data and compute the corresponding estimate of theta, say theta_tilde. A large set of tuples (phi, theta_tilde) can be generated in this manner. Nonparametric methods may be use to fit the function E(phi|theta_tilde=a), using these tuples. It is proposed to estimate phi using the fitted E(phi|theta_tilde=theta_hat), where theta_hat is the auxiliary estimate, using the real sample data. This is a consistent and asymptotically normally distributed estimator, under certain assumptions. Monte Carlo results for dynamic panel data and vector autoregressions show that this estimator can have very attractive small sample properties. Confidence intervals can be constructed using the quantiles of the phi for which theta_tilde is close to theta_hat. Such confidence intervals are found to have very accurate coverage.

Análisis de prestaciones de procesadores multi-core y multi-thread

Los procesadores multi-core y el multi-threading por hardware permiten aumentar el rendimiento de las aplicaciones. Por un lado, los procesadores multi-core combinan 2 o más procesadores en un mismo chip. Por otro lado, el multi-threading por hardware es una técnica que incrementa la utilización de los recursos del procesador. Este trabajo presenta un análisis de rendimiento de los resultados obtenidos en dos aplicaciones, multiplicación de matrices densas y transformada rápida de Fourier. Ambas aplicaciones se han ejecutado en arquitecturas multi-core que explotan el paralelismo a nivel de thread pero con un modelo de multi-threading diferente. Los resultados obtenidos muestran la importancia de entender y saber analizar el efecto del multi-core y multi-threading en el rendimiento.

A decision-making Fokker-Planck model in computational neuroscience

Minimal models for the explanation of decision-making in computational neuroscience are based on the analysis of the evolution for the average firing rates of two interacting neuron populations. While these models typically lead to multi-stable scenario for the basic derived dynamical systems, noise is an important feature of the model taking into account finite-size effects and robustness of the decisions. These stochastic dynamical systems can be analyzed by studying carefully their associated Fokker-Planck partial differential equation. In particular, we discuss the existence, positivity and uniqueness for the solution of the stationary equation, as well as for the time evolving problem. Moreover, we prove convergence of the solution to the the stationary state representing the probability distribution of finding the neuron families in each of the decision states characterized by their average firing rates. Finally, we propose a numerical scheme allowing for simulations performed on the Fokker-Planck equation which are in agreement with those obtained recently by a moment method applied to the stochastic differential system. Our approach leads to a more detailed analytical and numerical study of this decision-making model in computational neuroscience.

Large multi-gene phylogenetic trees of the grasses (Poaceae): progress towards complete tribal and generic level sampling.

In this paper we included a very broad representation of grass family diversity (84% of tribes and 42% of genera). Phylogenetic inference was based on three plastid DNA regions rbcL, matK and trnL-F, using maximum parsimony and Bayesian methods. Our results resolved most of the subfamily relationships within the major clades (BEP and PACCMAD), which had previously been unclear, such as, among others the: (i) BEP and PACCMAD sister relationship, (ii) composition of clades and the sister-relationship of Ehrhartoideae and Bambusoideae + Pooideae, (iii) paraphyly of tribe Bambuseae, (iv) position of Gynerium as sister to Panicoideae, (v) phylogenetic position of Micrairoideae. With the presence of a relatively large amount of missing data, we were able to increase taxon sampling substantially in our analyses from 107 to 295 taxa. However, bootstrap support and to a lesser extent Bayesian inference posterior probabilities were generally lower in analyses involving missing data than those not including them. We produced a fully resolved phylogenetic summary tree for the grass family at subfamily level and indicated the most likely relationships of all included tribes in our analysis.

Extended truncated Tweedie-Poisson model

It has been argued that by truncating the sample space of the negative binomial and of the inverse Gaussian-Poisson mixture models at zero, one is allowed to extend the parameter space of the model. Here that is proved to be the case for the more general three parameter Tweedie-Poisson mixture model. It is also proved that the distributions in the extended part of the parameter space are not the zero truncation of mixed poisson distributions and that, other than for the negative binomial, they are not mixtures of zero truncated Poisson distributions either. By extending the parameter space one can improve the fit when the frequency of one is larger and the right tail is heavier than is allowed by the unextended model. Considering the extended model also allows one to use the basic maximum likelihood based inference tools when parameter estimates fall in the extended part of the parameter space, and hence when the m.l.e. does not exist under the unextended model. This extended truncated Tweedie-Poisson model is proved to be useful in the analysis of words and species frequency count data.

Multimodel inference and multimodel averaging in empirical modeling of occupational exposure levels.

Empirical modeling of exposure levels has been popular for identifying exposure determinants in occupational hygiene. Traditional data-driven methods used to choose a model on which to base inferences have typically not accounted for the uncertainty linked to the process of selecting the final model. Several new approaches propose making statistical inferences from a set of plausible models rather than from a single model regarded as 'best'. This paper introduces the multimodel averaging approach described in the monograph by Burnham and Anderson. In their approach, a set of plausible models are defined a priori by taking into account the sample size and previous knowledge of variables influent on exposure levels. The Akaike information criterion is then calculated to evaluate the relative support of the data for each model, expressed as Akaike weight, to be interpreted as the probability of the model being the best approximating model given the model set. The model weights can then be used to rank models, quantify the evidence favoring one over another, perform multimodel prediction, estimate the relative influence of the potential predictors and estimate multimodel-averaged effects of determinants. The whole approach is illustrated with the analysis of a data set of 1500 volatile organic compound exposure levels collected by the Institute for work and health (Lausanne, Switzerland) over 20 years, each concentration having been divided by the relevant Swiss occupational exposure limit and log-transformed before analysis. Multimodel inference represents a promising procedure for modeling exposure levels that incorporates the notion that several models can be supported by the data and permits to evaluate to a certain extent model selection uncertainty, which is seldom mentioned in current practice.

Indirect likelihood inference

Given a sample from a fully specified parametric model, let Zn be a given finite-dimensional statistic - for example, an initial estimator or a set of sample moments. We propose to (re-)estimate the parameters of the model by maximizing the likelihood of Zn. We call this the maximum indirect likelihood (MIL) estimator. We also propose a computationally tractable Bayesian version of the estimator which we refer to as a Bayesian Indirect Likelihood (BIL) estimator. In most cases, the density of the statistic will be of unknown form, and we develop simulated versions of the MIL and BIL estimators. We show that the indirect likelihood estimators are consistent and asymptotically normally distributed, with the same asymptotic variance as that of the corresponding efficient two-step GMM estimator based on the same statistic. However, our likelihood-based estimators, by taking into account the full finite-sample distribution of the statistic, are higher order efficient relative to GMM-type estimators. Furthermore, in many cases they enjoy a bias reduction property similar to that of the indirect inference estimator. Monte Carlo results for a number of applications including dynamic and nonlinear panel data models, a structural auction model and two DSGE models show that the proposed estimators indeed have attractive finite sample properties.

Identifying genetic signatures of selection in a non-model species, alpine gentian (Gentiana nivalis L.), using a landscape genetic approach

It is generally accepted that most plant populations are locally adapted. Yet, understanding how environmental forces give rise to adaptive genetic variation is a challenge in conservation genetics and crucial to the preservation of species under rapidly changing climatic conditions. Environmental variation, phylogeographic history, and population demographic processes all contribute to spatially structured genetic variation, however few current models attempt to separate these confounding effects. To illustrate the benefits of using a spatially-explicit model for identifying potentially adaptive loci, we compared outlier locus detection methods with a recently-developed landscape genetic approach. We analyzed 157 loci from samples of the alpine herb Gentiana nivalis collected across the European Alps. Principle coordinates of neighbor matrices (PCNM), eigenvectors that quantify multi-scale spatial variation present in a data set, were incorporated into a landscape genetic approach relating AFLP frequencies with 23 environmental variables. Four major findings emerged. 1) Fifteen loci were significantly correlated with at least one predictor variable (R (adj) (2) > 0.5). 2) Models including PCNM variables identified eight more potentially adaptive loci than models run without spatial variables. 3) When compared to outlier detection methods, the landscape genetic approach detected four of the same loci plus 11 additional loci. 4) Temperature, precipitation, and solar radiation were the three major environmental factors driving potentially adaptive genetic variation in G. nivalis. Techniques presented in this paper offer an efficient method for identifying potentially adaptive genetic variation and associated environmental forces of selection, providing an important step forward for the conservation of non-model species under global change.

Profitability, uncertainty and multi-product firm product proliferation: The Spanish car industry

This article studies how product introduction decisions relate to profitability and uncertainty in the context of multi-product firms and product differentiation. These two features, common to many modern industries, have not received much attention in the literature as compared to the classical problem of firm entry, even if the determinants of firm and product entry are quite different. The theoretical predictions about the sign of the impact of uncertainty on product entry are not conclusive. Therefore, an econometric model relating firms’ product introduction decisions with profitability and profit uncertainty is proposed. Firm’s estimated profits are obtained from a structural model of product demand and supply, and uncertainty is proxied by profits’ variance. The empirical analysis is carried out using data on the Spanish car industry for the period 1990-2000. The results show a positive relationship between product introduction and profitability, and a negative one with respect to profit variability. Interestingly, the degree of uncertainty appears to be a driving force of entry stronger than profitability, suggesting that the product proliferation process in the Spanish car market may have been mainly a consequence of lower uncertainty rather than the result of having a more profitable market. Keywords: Product introduction, entry, uncertainty, multiproduct firms, automobile JEL codes: L11, L13

Quantitative genetic modeling and inference in the presence of nonignorable missing data.

Natural selection is typically exerted at some specific life stages. If natural selection takes place before a trait can be measured, using conventional models can cause wrong inference about population parameters. When the missing data process relates to the trait of interest, a valid inference requires explicit modeling of the missing process. We propose a joint modeling approach, a shared parameter model, to account for nonrandom missing data. It consists of an animal model for the phenotypic data and a logistic model for the missing process, linked by the additive genetic effects. A Bayesian approach is taken and inference is made using integrated nested Laplace approximations. From a simulation study we find that wrongly assuming that missing data are missing at random can result in severely biased estimates of additive genetic variance. Using real data from a wild population of Swiss barn owls Tyto alba, our model indicates that the missing individuals would display large black spots; and we conclude that genes affecting this trait are already under selection before it is expressed. Our model is a tool to correctly estimate the magnitude of both natural selection and additive genetic variance.

Steady-state brain glucose transport kinetics re-evaluated with a four-state conformational model.

Glucose supply from blood to brain occurs through facilitative transporter proteins. A near linear relation between brain and plasma glucose has been experimentally determined and described by a reversible model of enzyme kinetics. A conformational four-state exchange model accounting for trans-acceleration and asymmetry of the carrier was included in a recently developed multi-compartmental model of glucose transport. Based on this model, we demonstrate that brain glucose (G(brain)) as function of plasma glucose (G(plasma)) can be described by a single analytical equation namely comprising three kinetic compartments: blood, endothelial cells and brain. Transport was described by four parameters: apparent half saturation constant K(t), apparent maximum rate constant T(max), glucose consumption rate CMR(glc), and the iso-inhibition constant K(ii) that suggests G(brain) as inhibitor of the isomerisation of the unloaded carrier. Previous published data, where G(brain) was quantified as a function of plasma glucose by either biochemical methods or NMR spectroscopy, were used to determine the aforementioned kinetic parameters. Glucose transport was characterized by K(t) ranging from 1.5 to 3.5 mM, T(max)/CMR(glc) from 4.6 to 5.6, and K(ii) from 51 to 149 mM. It was noteworthy that K(t) was on the order of a few mM, as previously determined from the reversible model. The conformational four-state exchange model of glucose transport into the brain includes both efflux and transport inhibition by G(brain), predicting that G(brain) eventually approaches a maximum concentration. However, since K(ii) largely exceeds G(plasma), iso-inhibition is unlikely to be of substantial importance for plasma glucose below 25 mM. As a consequence, the reversible model can account for most experimental observations under euglycaemia and moderate cases of hypo- and hyperglycaemia.

Social inequalities and mortality in Europe-results from a large multi-national cohort

BACKGROUND Socio-economic inequalities in mortality are observed at the country level in both North America and Europe. The purpose of this work is to investigate the contribution of specific risk factors to social inequalities in cause-specific mortality using a large multi-country cohort of Europeans. METHODS A total of 3,456,689 person/years follow-up of the European Prospective Investigation into Cancer and Nutrition (EPIC) was analysed. Educational level of subjects coming from 9 European countries was recorded as proxy for socio-economic status (SES). Cox proportional hazard model's with a step-wise inclusion of explanatory variables were used to explore the association between SES and mortality; a Relative Index of Inequality (RII) was calculated as measure of relative inequality. RESULTS Total mortality among men with the highest education level is reduced by 43% compared to men with the lowest (HR 0.57, 95% C.I. 0.52-0.61); among women by 29% (HR 0.71, 95% C.I. 0.64-0.78). The risk reduction was attenuated by 7% in men and 3% in women by the introduction of smoking and to a lesser extent (2% in men and 3% in women) by introducing body mass index and additional explanatory variables (alcohol consumption, leisure physical activity, fruit and vegetable intake) (3% in men and 5% in women). Social inequalities were highly statistically significant for all causes of death examined in men. In women, social inequalities were less strong, but statistically significant for all causes of death except for cancer-related mortality and injuries. DISCUSSION In this European study, substantial social inequalities in mortality among European men and women which cannot be fully explained away by accounting for known common risk factors for chronic diseases are reported.

Applications of stochastic analysis in finance and statistical inference

First: A continuous-time version of Kyle's model (Kyle 1985), known as the Back's model (Back 1992), of asset pricing with asymmetric information, is studied. A larger class of price processes and of noise traders' processes are studied. The price process, as in Kyle's model, is allowed to depend on the path of the market order. The process of the noise traders' is an inhomogeneous Lévy process. Solutions are found by the Hamilton-Jacobi-Bellman equations. With the insider being risk-neutral, the price pressure is constant, and there is no equilibirium in the presence of jumps. If the insider is risk-averse, there is no equilibirium in the presence of either jumps or drifts. Also, it is analised when the release time is unknown. A general relation is established between the problem of finding an equilibrium and of enlargement of filtrations. Random announcement time is random is also considered. In such a case the market is not fully efficient and there exists equilibrium if the sensitivity of prices with respect to the global demand is time decreasing according with the distribution of the random time. Second: Power variations. it is considered, the asymptotic behavior of the power variation of processes of the form _integral_0^t u(s-)dS(s), where S_ is an alpha-stable process with index of stability 0&alpha&2 and the integral is an Itô integral. Stable convergence of corresponding fluctuations is established. These results provide statistical tools to infer the process u from discrete observations. Third: A bond market is studied where short rates r(t) evolve as an integral of g(t-s)sigma(s) with respect to W(ds), where g and sigma are deterministic and W is the stochastic Wiener measure. Processes of this type are particular cases of ambit processes. These processes are in general not of the semimartingale kind.

Induction of protective T-helper 1 immune responses against Echinococcus granulosus in mice by a multi-T-cell epitope antigen based on five proteins

In this study, we designed an experiment to predict a potential immunodominant T-cell epitope and evaluate the protectivity of this antigen in immunised mice. The T-cell epitopes of the candidate proteins (EgGST, EgA31, Eg95, EgTrp and P14-3-3) were detected using available web-based databases. The synthesised DNA was subcloned into the pET41a+ vector and expressed in Escherichia coli as a fusion to glutathione-S-transferase protein (GST). The resulting chimeric protein was then purified by affinity chromatography. Twenty female C57BL/6 mice were immunised with the antigen emulsified in Freund's adjuvant. Mouse splenocytes were then cultured in Dulbecco's Modified Eagle's Medium in the presence of the antigen. The production of interferon-γ was significantly higher in the immunised mice than in the control mice (> 1,300 pg/mL), but interleukin (IL)-10 and IL-4 production was not statistically different between the two groups. In a challenge study in which mice were infected with 500 live protoscolices, a high protectivity level (99.6%) was demonstrated in immunised BALB/C mice compared to the findings in the control groups [GST and adjuvant (Adj) ]. These results demonstrate the successful application of the predicted T-cell epitope in designing a vaccine against Echinococcus granulosus in a mouse model.