Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “model selection”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 865 records · Page 48Linked to original sources
PubMed2006

Sparse Bayesian kernel survival analysis for modeling the growth domain of microbial pathogens.

Survival analysis is a branch of statistics concerned with the time elapsing before "failure," with diverse applications in medical statistics and the analysis of the reliability of electrical or mechanical components. We introduce a parametric accelerated life survival analysis model based on kernel learning methods that, at least in principal, is able to learn arbitrary dependencies between a vector of explanatory variables and the scale of the distribution of survival times. The proposed kernel survival analysis method is then used to model the growth domain of Clostridium botulinum, the food processing and storage conditions permitting the growth of this foodborne microbial pathogen, leading to the production of the neurotoxin responsible for botulism. A Bayesian training procedure, based on the evidence framework, is used for model selection and to provide a credible interval on model predictions. The kernel survival analysis models are found to be more accurate than models based on more traditional survival analysis techniques but also suggest a risk assessment of the foodborne botulism hazard would benefit from the collection of additional data.

Artificial Intelligence↗
PubMed2000

Differential migration from high fitness demes in the shining fungus beetle, Phalacrus substriatus.

Using data from three years (1994-1996), I tested whether differential migration occurs from demes of high mean fitness in the shining fungus beetle, Phalacrus substriatus. The results show evidence for differential migration, thus providing evidence from a natural population for a critical demographic assumption of many interdemic selection models. To predict the evolutionary response to interdemic selection through differential migration, the genetic basis of the variation among demes in mean fitness must be known because the observed patterns could also be explained by some demes having an intrinsically favorable habitat. Thus, the importance of differential migration through interdemic selection in natural populations cannot be unequivocally answered without experiments specifically addressing the question of what causes differences in mean fitness among demes.

Animals↗
PubMed2006

Quantifying uncertainty in the potential distribution of an invasive species: climate and the Argentine ant.

Maps of a species' potential range make an important contribution to conservation and invasive species risk analysis. Spatial predictions, however, should be accompanied by an assessment of their uncertainty. Here, we use multimodel inference to generate confidence intervals that incorporate both the uncertainty involved in model selection as well as the error associated with model fitting. In the case of the invasive Argentine ant, we found that it was most likely to occur where the mean daily temperature in mid-winter was 7-14 degrees C and maximum daily temperatures during the hottest month averaged 19-30 degrees C. Uninvaded regions vulnerable to future establishment include: southern China, Taiwan, Zimbabwe, central Madagascar, Morocco, high-elevation Ethiopia, Yemen and a number of oceanic islands. Greatest uncertainty exists over predictions for China, north-east India, Angola, Bolivia, Lord Howe Island and New Caledonia. Quantifying the costs of different errors (false negatives vs. false positives) was considered central for connecting modelling to decision-making and management processes.

Animals↗
PubMed1992

Selectivity of a CaCl2 continuous infusion screening method in rats.

In albino rats the combined use of two methods based on a continuous perfusion of aconitine nitrate (0.15 mg/ml) or calcium chloride (0.4 M) demonstrated the selectivity of the calcium chloride-induced dysrhythmias model. This rapid screening was carried out with antidysrhythmic agents according to the Vaughan-Williams's classification (1): quinidine (class I) 5 mg/kg; atenolol (class II) 2 mg/kg; amiodarone (class III) 10 mg/kg and verapamil (class IV) 2 mg/kg. A compound synthesized in our laboratory was used to verify model selectivity.

Aconitine↗
PubMed1977

Regions of stable equilibria for models of differential selection in the two sexes under random mating.

The equilibrium structure of models of differential selection in the sexes is investigated. It is shown that opposing additive selection leads to stable polymorphic equilibria for only a restricted set of selection intensities, and that for weak selection intensities must be of approximately the same magnitude in the sexes. General models of opposing directional selection, with arbitrary dominance, are investigated by considering simultaneously the stability properties of the trivial equilibria and the curve along which multiple roots appear. Numerical calculations lead us to infer that the average degree of dominance determines the equilibrium characteristics of models of opposing selection. It appears that if the favored alleles are, on the average, recessive, there may be multiple polymorphic equilibria, whereas only a single polymorphic equilibrium can occur when the favored alleles are, on the average, dominant. The principle that the average degree of dominance controls equilibrium behavior is then extended to models allowing directional selection in one sex with overdominance in the other sex, by showing that polymorphism is maintained if and only if the average fitness in heterozygotes exceeds one.

Animals↗
PubMed1994

Children's integration of speech and pointing gestures in comprehension.

We examined 4- and 9-year-old's referential comprehension when given pointing gestures and spoken labels, in two types of contextually ambiguous situations. In one situation, speech/gesture discordance was produced in conditions where labels for one of four objects being referred to sounded either alike, or different from each other. In the other, the contextual set contained the same two objects, and ambiguity was produced by factorially combining speech on a continuum ranging between /bcl/ and /dcl/ with a pointing gesture from a continuum ranging between an unambiguous point to a ball and to a doll. Results showed that the speech modality had a far greater influence on word comprehension than gestures. Second, the influence of gestures was greater for the older children. Mathematical models of speech-gesture understanding were tested against the data. Selection models assume that one dimension of information is used on a given trial, and that the selection of a modality depends on the ambiguity of information encoded on the dominant dimension. The Fuzzy Logical Model of Perception (FLMP) assumes that both modalities are evaluated independently of one another and then integrated to achieve comprehension. The results from both age groups were best described by the assumptions of the FLMP. Results are related to general claims about perceptual development during childhood concerning the quality of representations formed and dimensional selectivity of visual-spoken language.

Auditory Perception↗
PubMed1999

Computational models of coherent and transparent plaid motion.

The perceived motion of two added sinusoidal gratings of similar amplitude and spatial frequency but different orientations is often coherent. However, when either relative grating contrast or frequency are varied, perception may transform to a motion transparency. For plaids, both multiplicative and additive transparent percepts are reported. To explain perception, several computational models of motion transparency are proposed. The most general model considered is, however, a quadratic form with five unknowns. To stabilize the transparent model, additional constraints are introduced so that two velocities may be detected from the motion of plaid patterns. It is shown how this model may be realised by a two-layer (linear) feedforward network and how network learning paradigms may be used to explain some facets of visual perception. To describe the motion of plaid patterns there is an ambiguity because computational models of both coherent and transparent motion may be used to detect image velocity. In view of this competition between models, the issue of model selection is addressed; especially for cases where two or more models fit the image measurements without a residual error. The computational approach that is proposed affords one explanation why perception selects transparency in favour of coherence for plaid patterns by adjustments of relative grating contrast and frequency.

Humans↗
PubMed1999

Ultrastructural model for size selectivity in glomerular filtration.

A theoretical model was developed to relate the size selectivity of the glomerular barrier to the structural characteristics of the individual layers of the capillary wall. Thicknesses and other linear dimensions were evaluated, where possible, from previous electron microscopic studies. The glomerular basement membrane (GBM) was represented as a homogeneous material characterized by a Darcy permeability and by size-dependent hindrance coefficients for diffusion and convection, respectively; those coefficients were estimated from recent data obtained with isolated rat GBM. The filtration slit diaphragm was modeled as a single row of cylindrical fibers of equal radius but nonuniform spacing. The resistances of the remainder of the slit channel, and of the endothelial fenestrae, to macromolecule movement were calculated to be negligible. The slit diaphragm was found to be the most restrictive part of the barrier. Because of that, macromolecule concentrations in the GBM increased, rather than decreased, in the direction of flow. Thus the overall sieving coefficient (ratio of Bowman's space concentration to that in plasma) was predicted to be larger for the intact capillary wall than for a hypothetical structure with no GBM. In other words, because the slit diaphragm and GBM do not act independently, the overall sieving coefficient is not simply the product of those for GBM alone and the slit diaphragm alone. Whereas the calculated sieving coefficients were sensitive to the structural features of the slit diaphragm and to the GBM hindrance coefficients, variations in GBM thickness or filtration slit frequency were predicted to have little effect. The ability of the ultrastructural model to represent fractional clearance data in vivo was at least equal to that of conventional pore models with the same number of adjustable parameters. The main strength of the present approach, however, is that it provides a framework for relating structural findings to the size selectivity of the glomerular barrier.

Animals↗
PubMed2005

[Dynamic regimes in a model of single-locus density-dependent selection].

A model of density-dependent selection in a Mendelian single-locus population was analyzed in the case where the fitnesses of genotypic forms are exponential functions of the population size. Analytical and numerical studies of the model were performed for a diallelic locus, and parametric regions were established for different dynamic behaviors of the model. The diallelic model of density-dependent selection was generalized to a multiallelic locus; the results of its analysis are described.

Models, Genetic↗
PubMed1990

On the locus of visual selection: evidence from focused attention tasks.

Early and late selection models of attention disagree about whether visual objects are identified before or after selection, with recent evidence of interference from to-be-ignored stimuli favoring late selection over early selection accounts. However, these tests may not have permitted optimal attentional focusing. In 4 experiments subjects identified an attentionally cued target letter embedded among distractors. Only minimal effects of information appearing in to-be-ignored locations were observed. This striking efficiency of selection provides support for early selection theories and calls into question some late selection theories holding that stimuli throughout the display are immediately and fully identified prior to attentional selection. In order to explain the larger pattern of results across a variety of focused- and divided-attention paradigms, a hybrid model is advanced with a flexible locus for visual selection.

Adult↗
PubMed1985

In vivo evidence for the selectivity of ICI 154129 for the delta-opioid receptor.

The in vivo selectivity of the novel delta opioid-receptor antagonist N,N-bisallyl-Tyr-Gly-Gly-psi-(CH2S)-Phe-Leu-OH (ICI 154129) was examined in several opioid-selective models. Antagonism at the delta receptor was demonstrated in the striatal head-turn model in the rat. Intrapallidal injection of the relatively selective delta-receptor agonist D-Ala2,D-Leu5-enkephalin (0.5 micrograms) slowed the head-turn time and this effect was completely prevented by prior subcutaneous administration of ICI 154129 (30 mg/kg). The role of delta receptors in two classical test situations was studied using the mixed opioid agonist etorphine and the antagonists naloxone and ICI 154129. The drug ICI 154129 (30 mg/kg, s.c.) failed to prevent the antinociceptive effects and stimulation of locomotor activity produced by etorphine, whereas the relatively selective mu-opioid receptor antagonist, naloxone was effective in both test situations. The possible involvement of delta receptors in morphine-induced dependence was studied by monitoring the abstinence behaviour precipitated in rats given pellets of morphine by either ICI 154129 or naloxone. Naloxone (0.5 mg/kg, i.p.) precipitated a characteristic withdrawal syndrome in conscious rats and, at a much smaller dose (0.02 mg/kg, i.p.), induced shaking behaviour in pentobarbitone-anaesthetised rats. No withdrawal signs were observed in either model after injection of ICI 154129 (30 mg/kg, s.c.), suggesting that the delta receptors are not involved in dependence on morphine.

Analgesics↗
PubMed1995

Perturbation analysis of a two-locus model with directional selection and recombination.

A population genetic two-locus model with additive, directional selection and recombination is considered. It is assumed that recombination is weaker than selection; i.e., the recombination parameter r is smaller than the selection coefficients. This assumption is appropriate for describing the effects of two-locus selection at the molecular level. The model is formulated in terms of ordinary differential equations (ODEs) for the gamete frequencies x = (x1, x2, x3, x4), defined on the simplex S4. The ODEs are analyzed using first a regular perturbation technique. However, this approach yields satisfactory results only if r is very small relative to the selection coefficients and if the initial values x(0) are in the interior part of S4. To cope with this problem, a novel two-scale perturbation method is proposed which rests on the theory of averaging of vectorfields. It is demonstrated that the zeroth-order solution of this two-scale approach approximates the numerical solution of the model well, even if recombination rate is on the order of the selection coefficients.

Animals↗
PubMed1996

Variational principles in evolution.

For a one-locus selection model, Svirezhev introduced an integral variational principle by defining a Lagrangian which remained stationary on the trajectory followed by the population undergoing selection. It is shown here (i) that this principle can be extended to multiple loci in some simple cases and (ii) that the Lagrangian is defined by a straightforward generalization of the one-locus case, but (iii) that in two-locus or more general models there is no straightforward extension of this principle if linkage and epistasis are present. The population trajectories can be constructed as trajectories of steepest ascent in a Riemannian metric space. A general method is formulated to find the metric tensor and the surface in the metric space on which the trajectories, which characterize the variations in the gene structure of the population, lie. The local optimality principle holds good in such a space. In the special case when all possible linkage disequilibria are zero, the phase point of the n-locus genetic system moves on the surface of the product space of n higher dimensional unit spheres in a certain Riemannian metric space of gene frequencies so that the rate of change of mean fitness is maximum along the trajectory. In the two-locus case the corresponding surface is a hyper-torus.

Biological Evolution↗
PubMed1997

Selective processing of superimposed objects: an electrophysiological analysis of object-based attentional selection.

We investigated whether object-based attentional selection occurs from grouped-array or spatially-invariant representations. Subjects were presented with colored objects and asked to judge whether a particular color/shape conjunction was present, regardless of whether the color and shape were part of a single object (same-object condition) or occurred on two different objects (different-object condition). RTs and accuracies were recorded for subjects judgments. ERP components, in particular the P1 and N1, were elicited both from the presentation of the target objects and from a post-display probe that was employed as an index of spatial attention. Consistent with predictions of object-based selection models, RTs and accuracies were faster on same than on different object trials. N1s elicited by the target objects and P1s elicited by the post-display probes discriminated between same and different object trials when the two target objects were superimposed. These data are consistent with the proposal that object-based selection is spatially mediated, even in the case of partially overlapping objects. The data are discussed in terms of space- and object-based models of visual selective attention.

Adolescent↗
PubMed2005

Genetic variation and selection response in model breeding populations of Brassica rapa following a diversity bottleneck.

Domestication and breeding share a common feature of population bottlenecks followed by significant genetic gain. To date, no crop models for investigating the evolution of genetic variance, selection response, and population diversity following bottlenecks have been developed. We developed a model artificial selection system in the laboratory using rapid-cycling Brassica rapa. Responses to 10 cycles of recurrent selection for cotyledon size were compared across a broad population founded with 200 individuals, three bottleneck populations initiated with two individuals each, and unselected controls. Additive genetic variance and heritability were significantly larger in the bottleneck populations prior to selection and this corresponded to a heightened response of bottleneck populations during the first three cycles. However, the overall response was ultimately greater and more sustained in the broad population. AFLP marker analyses revealed the pattern and extent of population subdivision were unaffected by a bottleneck even though the diversity retained in a selection population was significantly limited. Rapid gain in genetically more uniform bottlenecked populations, particularly in the short term, may offer an explanation for why domesticators and breeders have realized significant selection progress over relatively short time periods.

Brassica rapa↗
PubMed2006

Likelihood ratio tests in behavioral genetics: problems and solutions.

The likelihood ratio test of nested models for family data plays an important role in the assessment of genetic and environmental influences on the variation in traits. The test is routinely based on the assumption that the test statistic follows a chi-square distribution under the null, with the number of restricted parameters as degrees of freedom. However, tests of variance components constrained to be non-negative correspond to tests of parameters on the boundary of the parameter space. In this situation the standard test procedure provides too large p-values and the use of the Akaike Information Criterion (AIC) or the Bayesian Information Criterion (BIC) for model selection is problematic. Focusing on the classical ACE twin model for univariate traits, we adapt existing theory to show that the asymptotic distribution for the likelihood ratio statistic is a mixture of chi-square distributions, and we derive the mixing probabilities. We conclude that when testing the AE or the CE model against the ACE model, the p-values obtained from using the chi(2)(1 df) as the reference distribution should be halved. When the E model is tested against the ACE model, a mixture of chi(2)(0 df), chi(2)(1 df) and chi(2)(2 df) should be used as the reference distribution, and we provide a simple formula to compute the mixing probabilities. Similar results for tests of the AE, DE and E models against the ADE model are also derived. Failing to use the appropriate reference distribution can lead to invalid conclusions.

Analysis of Variance↗
PubMed2005

Constructing a speculative kernel machine for pattern classification.

We propose and investigate the performance of a new geometry-based algorithm designed to identify potentially informative data points for classification. An incremental QR update scheme is used to build a classifier using a subset of these points as radial basis function centers. The minimum descriptive length and the leave-one-out error criteria are employed for automatic model selection. The proposed scheme is shown to generate parsimonious models, which perform generalization comparable to the state-of-the-art support and relevance vector machines.

Algorithms↗
PubMed2004

Likelihood-based inference for stochastic models of sexual network formation.

Sexually-transmitted diseases (STDs) constitute a major public health concern. Mathematical models for the transmission dynamics of STDs indicate that heterogeneity in sexual activity level allow them to persist even when the typical behavior of the population would not support endemicity. This insight focuses attention on the distribution of sexual activity level in a population. In this paper, we develop several stochastic process models for the formation of sexual partnership networks. Using likelihood-based model selection procedures, we assess the fit of the different models to three large distributions of sexual partner counts: (1) Rakai, Uganda, (2) Sweden, and (3) the USA. Five of the six single-sex networks were fit best by the negative binomial model. The American women's network was best fit by a power-law model, the Yule. For most networks, several competing models fit approximately equally well. These results suggest three conclusions: (1) no single unitary process clearly underlies the formation of these sexual networks, (2) behavioral heterogeneity plays an essential role in network structure, (3) substantial model uncertainty exists for sexual network degree distributions. Behavioral research focused on the mechanisms of partnership formation will play an essential role in specifying the best model for empirical degree distributions. We discuss the limitations of inferences from such data, and the utility of degree-based epidemiological models more generally.

Databases as Topic↗