Home About us Contact
|
Home About us Contact | ||
|
|
||
Selected Model (selected + model)
Selected AbstractsStudy of the space,time effects in the concentration of airborne pollutants in the Metropolitan Region of Rio de JaneiroENVIRONMETRICS, Issue 4 2003Marina Silva Paez Abstract In this article, we present an application of models with temporal and spatial components, from the Bayesian point of view, on data pollutants collected in 16 different monitoring sites located in the Metropolitan Area of Rio de Janeiro during 1999. All the models considered here assume conditionally independent observations, with a mean specified by the sum of random temporal and spatial components and a linear function of the maximum daily temperature and indicators of the day of the week. Our aim here is to analyze distinct specifications for the components, assuming different kinds of modeling that are not usually compared. The comparison is based on the posterior predictive loss function proposed by Gelfand and Ghosh (1998). The best specifications for the spatial component were the ones which considered a geostatistical approach to its correlation function. The best specification for the temporal component was the stationary autoregressive form. The pollutant concentrations were interpolated in a grid of points in the area of higher population density at a fixed period of time for the selected model. Copyright © 2003 John Wiley & Sons, Ltd. [source] Factors influencing territorial occupancy and reproductive output in the Booted Eagle Hieraaetus pennatusIBIS, Issue 4 2006JOSÉ E. MARTÍNEZ During a 7-year research project in a forested area of southeastern Spain, we studied territorial occupancy and reproductive success in a Booted Eagle Hieraaetus pennatus population. We monitored 65 territories, gathering information on 406 occupancy events and 229 breeding attempts, including those of two potential competitors, the Northern Goshawk Accipiter gentilis and the Common Buzzard Buteo buteo. Generalized linear mixed models were used to explain occupancy and productivity, by evaluating the relative contribution of three different types of variables (habitat, competition and past events) and considering territory as a random effect. We examined a set of a priori hypothesized models, together with a number of additional models, and selected the best models following an information-theoretic approach. Our best models related territorial occupancy and productivity to previous breeding success (the fledging of one or two young), which appeared to be the most important factor determining the probability of reoccupation and the reproductive output in the subsequent year. The best occupation model revealed that the probabilities of occupancy were also conditioned by a competition variable (intraspecific nearest-neighbour distance) and two habitat variables (the location of the nest on the valley slope and the distance to the nearest forest track). Unlike the best occupation model, however, the selected model for reproductive output did not incorporate any competition variable besides previous breeding success, but included another two habitat variables (the effects of trunk height and NNE orientation). [source] Analysis and selection criteria of BSIM4 flicker noise simulation modelsINTERNATIONAL JOURNAL OF CIRCUIT THEORY AND APPLICATIONS, Issue 7 2008T. Noulis Abstract CMOS transistors' noise performance is mainly dominated by flicker (1/f) noise. BSIM4.X MOSFET simulation model develops two distinct models, SPICE-Flicker and BSIM-Flicker, to calculate flicker noise. In this paper, these two models are analytically examined and compared to noise measurements, using an NMOS and a PMOS device fabricated in 0.6µm process by Austria Mikro Systeme (AMS). MOSFET 1/f noise measurements and the respective simulations were obtained under various bias conditions, as to study which flicker noise model is the optimum in each operating region. Measurement temperature was constant at 295,K. Comparisons suggest that in an NMOS transistor operating in the triode or saturation region, BSIM-Flicker model is accurate and therefore preferable. In a PMOS transistor, the most suitable model to describe its 1/f noise performance in the linear regime is also BSIM-Flicker, whereas SPICE-Flicker is more preferable in saturation. In NMOS transistors, the selected model provides a great accurate description of flicker noise, contrary to PMOS transistors, where simulation models appear to be quite unreliable and need further improvement. Copyright © 2007 John Wiley & Sons, Ltd. [source] Related-variables selection in temporal disaggregationJOURNAL OF FORECASTING, Issue 4 2009Kosei Fukuda Abstract Two related-variables selection methods for temporal disaggregation are proposed. In the first method, the hypothesis tests for a common feature (cointegration or serial correlation) are first performed. If there is a common feature between observed aggregated series and related variables, the conventional Chow,Lin procedure is applied. In the second method, alternative Chow,Lin disaggregating models with and without related variables are first estimated and the corresponding values of the Bayesian information criterion (BIC) are stored. It is determined on the basis of the selected model whether related variables should be included in the Chow,Lin model. The efficacy of these methods is examined via simulations and empirical applications. Copyright © 2008 John Wiley & Sons, Ltd. [source] Can out-of-sample forecast comparisons help prevent overfitting?JOURNAL OF FORECASTING, Issue 2 2004Todd E. ClarkArticle first published online: 3 MAR 200 Abstract This paper shows that out-of-sample forecast comparisons can help prevent data mining-induced overfitting. The basic results are drawn from simulations of a simple Monte Carlo design and a real data-based design similar to those used in some previous studies. In each simulation, a general-to-specific procedure is used to arrive at a model. If the selected specification includes any of the candidate explanatory variables, forecasts from the model are compared to forecasts from a benchmark model that is nested within the selected model. In particular, the competing forecasts are tested for equal MSE and encompassing. The simulations indicate most of the post-sample tests are roughly correctly sized. Moreover, the tests have relatively good power, although some are consistently more powerful than others. The paper concludes with an application, modelling quarterly US inflation. Copyright © 2004 John Wiley & Sons, Ltd. [source] Fat-tailed gene flow in the dioecious canopy tree species Fraxinus mandshurica var. japonica revealed by microsatellitesMOLECULAR ECOLOGY, Issue 10 2006S. GOTO Abstract Pollen flow, seed dispersal and individual reproductive success can be simultaneously estimated from the genotypes of adults and offspring using stochastic models. Using four polymorphic microsatellite loci, gene flow of the wind-pollinated and wind-seed-dispersed dioecious tree species, Fraxinus mandshurica var. japonica, was quantified in a riparian forest, in northern Japan. In a 10.5-ha plot, 74 female adults, 76 male adults and 292 current-year seedlings were mapped and genotyped, together with 200 seeds. To estimate dispersal kernels of pollen and seeds, we applied normal, exponential power, Weibull, bivariate t -distribution kernels, and two-component models consisting of two normal distribution functions, one with a small and one with a large variance. A two-component pollen flow model with a small contribution (26.1%) from short-distance dispersal (, = 7.2 m), and the rest from long-distance flow (, = 209.9 m), was chosen for the best-fitting model. The average distance that integrated pollen flows inside and outside the study plot was estimated to be 196.8 m. Tree size and flowering intensity affected reproduction, and there appeared to be critical values that distinguished reproductively successful and unsuccessful adults. In contrast, the gene flow model that estimated both pollen and seed dispersal from established seedlings resulted in extensive seed dispersal, and the expected spatial genetic structures did not satisfactorily fit with the observations, even for the selected model. Our results advanced small-scale individual-based parentage analysis for quantifying fat-tailed gene flow in wind-mediated species, but also clarified its limitations and suggested future possibilities for gene flow studies. [source] UPPER BOUNDS ON THE MINIMUM COVERAGE PROBABILITY OF CONFIDENCE INTERVALS IN REGRESSION AFTER MODEL SELECTIONAUSTRALIAN & NEW ZEALAND JOURNAL OF STATISTICS, Issue 3 2009Paul Kabaila Summary We consider a linear regression model, with the parameter of interest a specified linear combination of the components of the regression parameter vector. We suppose that, as a first step, a data-based model selection (e.g. by preliminary hypothesis tests or minimizing the Akaike information criterion , AIC) is used to select a model. It is common statistical practice to then construct a confidence interval for the parameter of interest, based on the assumption that the selected model had been given to us,a priori. This assumption is false, and it can lead to a confidence interval with poor coverage properties. We provide an easily computed finite-sample upper bound (calculated by repeated numerical evaluation of a double integral) to the minimum coverage probability of this confidence interval. This bound applies for model selection by any of the following methods: minimum AIC, minimum Bayesian information criterion (BIC), maximum adjusted,R2, minimum Mallows' CP and,t -tests. The importance of this upper bound is that it delineates general categories of design matrices and model selection procedures for which this confidence interval has poor coverage properties. This upper bound is shown to be a finite-sample analogue of an earlier large-sample upper bound due to Kabaila and Leeb. [source] Variable Selection for Logistic Regression Using a Prediction-Focused Information CriterionBIOMETRICS, Issue 4 2006Gerda Claeskens Summary In biostatistical practice, it is common to use information criteria as a guide for model selection. We propose new versions of the focused information criterion (FIC) for variable selection in logistic regression. The FIC gives, depending on the quantity to be estimated, possibly different sets of selected variables. The standard version of the FIC measures the mean squared error of the estimator of the quantity of interest in the selected model. In this article, we propose more general versions of the FIC, allowing other risk measures such as the one based on Lp error. When prediction of an event is important, as is often the case in medical applications, we construct an FIC using the error rate as a natural risk measure. The advantages of using an information criterion which depends on both the quantity of interest and the selected risk measure are illustrated by means of a simulation study and application to a study on diabetic retinopathy. [source] | |||||||||||||