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 811 records · Page 45Linked to original sources
PubMed1982

Loneliness as a function of selected personality variables.

Hypothesized that in a multivariate statistical model, selected personality variables, depression, anxiety, neuroticism, psychoticism, misanthropy, and external locus of control, could positively predict loneliness, and self-esteem and extraversion could negatively predict loneliness scores. Two groups of Ss were studied independently. Ss in Group I were 232 Iranian college students (156 males, 76 females) who were studying in American colleges. Group II consisted of 305 Iranian students (168 males, 137 females) who were studying in Iranian universities. The obtained results, applying multiple regression analysis, confirmed the directions stated in the research hypothesis. However, some of the selected variables did not contribute significantly in the regression equations. Because of fluctuation of the regression coefficients, due to multicolinearity, the data were subjected to factor analysis. Two factors emerged with eigenvalues greater than unity. Loneliness loaded heavily on the first factor, which was identified as negative attribute of personality.

Adult↗
PubMed2004

A maximum principle for the mutation-selection equilibrium of nucleotide sequences.

We study the equilibrium behaviour of a deterministic four-state mutation-selection model as a model for the evolution of a population of nucleotide sequences in sequence space. The mutation model is the Kimura 3ST mutation scheme, and the selection scheme is assumed to be invariant under permutation of sites. Considering the evolution process both forward and backward in time, we use the ancestral distribution as the stationary state of the backward process to derive an expression for the mutational loss (as the difference between ancestral and population mean fitness), and we prove a maximum principle that determines the population mean fitness in mutation-selection balance.

Animals↗
PubMed2001

Object segmentation model: analytical results and biological implications.

A simple, biologically motivated neural network for segmentation of a moving object from a visual scene is presented. The model consists of two parts: an object selection model which employs a scaling approach for receptive field sizes, and a subsequent network implementing a spotlight by means of multiplicative synapses. The network selects one object out of several, segments the rough contour of the object, and encodes the winner object's position with high accuracy. Analytical equations for the performance level of the network, e.g., for the critical distance of two objects above which they are perceived as separate, are derived. The network preferentially chooses the object with the largest angular velocity and the largest angular width. An equation for the velocity and width preferences is presented. Additionally it is shown that for certain neurons of the model, flat receptive fields are more favourable than Gaussian ones. The network exhibits performances similar to those known from amphibians. Various electrophysiological and behavioral results--e.g., the distribution of the diameters of the receptive fields of tectal neurons, of the tongue-projecting salamander Hydromantes italicus and the range of optimal prey velocities for prey catching--can be understood on the basis of the model.

Animals↗
PubMed2005

Applications of beta-mixture models in bioinformatics.

SUMMARY: We propose a beta-mixture model approach to solve a variety of problems related to correlations of gene-expression levels. For example, in meta-analyses of microarray gene-expression datasets, a threshold value of correlation coefficients for gene-expression levels is used to decide whether gene-expression levels are strongly correlated across studies. Ad hoc threshold values such as 0.5 are often used. In this paper, we use a beta-mixture model approach to divide the correlation coefficients into several populations so that the large correlation coefficients can be identified. Another important application of the proposed method is in finding co-expressed genes. Two examples are provided to illustrate both applications. Through our analysis, we also discover that the popular model selection criteria BIC and AIC are not suitable for the beta-mixture model. To determine the number of components in the mixture model, we suggest an alternative criterion, ICL-BIC, which is shown to perform better in selecting the correct mixture model. SUPPLEMENTARY INFORMATION: http://odin.mdacc.tmc.edu/~yuanj/highcorgeneanno.html.

Animals↗
PubMed2005

Variable selection for marginal longitudinal generalized linear models.

Variable selection is an essential part of any statistical analysis and yet has been somewhat neglected in the context of longitudinal data analysis. In this article, we propose a generalized version of Mallows's C(p) (GC(p)) suitable for use with both parametric and nonparametric models. GC(p) provides an estimate of a measure of model's adequacy for prediction. We examine its performance with popular marginal longitudinal models (fitted using GEE) and contrast results with what is typically done in practice: variable selection based on Wald-type or score-type tests. An application to real data further demonstrates the merits of our approach while at the same time emphasizing some important robust features inherent to GC(p).

Bias↗
PubMed1976

Modification of self-disclosing behaviours through modeling and vicarious reinforcement.

Modeling techniques were used to model moderate disclosure to high and low disclosers. Eighty female undergraduates (40 low disclosers and 40 high disclosers) listened to a model select items that were rated as moderate disclosures for discussion. Two modeling conditions were employed; in one condition the model was reinforced by the E (vicarious reinforcement), in the second condition the model was not reinforced. Two control conditions, one with the model present, the other with the model absent were also included. The results of two separate 2 X 4 X 2 repeated measures, hierarchical analyses of variance on the subject distance from modeled intimacy, and the number of moderate items selected on pre- and posttests, indicated that both modeling conditions were successful in producing moderation of self-disclosure.

Behavior Therapy↗
PubMed1994

Origin of a T cell clone with a mismatched combination of MHC restriction and coreceptor expression.

Although the existence of a large number of CD8+ class II MHC-specific CTLs had long been noticed, the origin of such T cells with a discordant combination of specificity and phenotype has been a mystery in the positive selection model. Recent reports suggesting the independency of the positive selection of T cells from coreceptor-mediated signals raised a possibility that they might be the progeny of putative transitional, mismatched, single-positive cells appearing before positive selection as proposed in the stochastic/selective model. By developing transgenic mice carrying TCR alpha and beta chain genes of a CD8+ class II MHC Ag-specific allogeneic CTL clone QM11, the origin of such T cells with mismatched TCR specificity and coreceptor expression was studied. The results indicate that QM11 belongs to a conventional CD8+ T cell population whose maturation is dependent on a class I (or class I-like) MHC product. Consequently, the reactivity of QM11 to I-Ak can be considered to be an accidental cross-reaction.

Animals↗
PubMed1982

Covariance analyses with heterogeneity of slopes in fixed models.

Techniques that describe the use of covariance when heterogeneity of slopes exists are severely limited. Although a few procedures for model selection have been recommended, none, except the hierarchical approach, is straightforward and usable with present computer programs. The hierarchical subset selection procedure presented in this paper is based on the proposition that heterogeneity may be present only for certain terms in the model. After hierarchical selection, those terms which do not involve heterogeneity are interpreted as in the usual analysis for covariance. The interpretations of those terms which do involve heterogeneity are modified with respect to significance tests performed at various values of the covariate. The hierarchical subset selection method allows one to investigate heterogeneity of slopes in covariance models as functions of the classification variables present in the design.

Analysis of Variance↗
PubMed2004

Multilevel selection and the partitioning of covariance: a comparison of three approaches.

Where the evolution of a trait is affected by selection at more than one hierarchical level, it is often useful to compare the magnitude of selection at each level by asking how much of the total evolutionary change is attributable to each level of selection. Three statistical partitioning techniques, each designed to answer this question, are compared, in relation to a simple multilevel selection model in which a trait's evolution is affected by both individual and group selection. None of the three techniques is wholly satisfactory: one implies that group selection can operate even if individual fitness is determined by individual phenotype alone, whereas the other two imply that group selection can operate even if there is no variance in group fitness. This has significant implications both for our understanding of what the term "multilevel selection" means and for the traditional concept of group selection.

Altruism↗
PubMed2005

[Research advance on lake ecosystem dynamic models].

Starting with the role of system analysis in lake ecosystem research, this paper summarized the tentative procedures and softwares for studying the dynamics of lake ecosystem. There are several main stages in modeling the dynamics of lake ecosystem, namely, problems identification, mathematical formulation, computation, validation, sensitive analysis, calibration, and verification. In the modeling, selecting temporal and spatial scales is essential but complex. Since 1960s, a rapid progress has been made in modeling the dynamics of lake ecosystem, being developed from simple zero-dimension models to complex ecological-aquatic-hydrodynamic ones, among which, exergy was applied popularly as an objective function in modeling. In this paper, LakeWeb and LEEDS (Lake Eutrophication, Effect, Dose, and Sensitivity model) were analyzed as examples. In China, the development of lake ecosystem dynamic models could be traced back to 1980s, and most of them were focused on Lake Dianch, Lake Taihu, Lake Chaohu and Lake Donghu. Some softwares such as CE-QUAL-ICM, WASP, AQUATOX, PAMOLARE and CAEDYM were developed to simulate lake ecosystem dynamics, among which, CE-QUAL-ICM is more suitable for long and narrow water bodies. WASP consists of three parts, i. e., DYNHYD, EUTRO, and TOXI. AQUATOX is an ecological risk model, and the parameters are mainly calibrated in U. S. A, which has limited its further application in China. The software ECOPATH for simulating the energy flows in lakes was also described in this paper. There are still many shortages in the lake ecosystem dynamic models, e. g., the lack of sufficient monitoring data for validation, insufficient consideration of uncertainties and the role of bacteria, and inconsistent relationship with watershed changes. The uncertainties are mainly from the intrinsic uncertainties in aquatic ecosystem, in modeling, in parameters selection, and also in forecast and application. Setting up long-term monitoring and data sharing mechanism, using interpolation to make data more densely, introducing objective functions, dealing with uncertainties, and constructing watershed-lake ecosystem dynamic model could be the available ways for overcoming the shortages.

China↗
PubMed1992

Model building of DNA/RNA triple helix containing L-deoxyadenosine.

A three-dimensional model of DNA/RNA triple helix that contains a poly(L-deoxyadenosine) (L-dA) chain is proposed based on computer-assisted model building and energy calculations. The model building was performed by a new method that systematically searches possible conformations of nucleotide units in the helical chains. Two possible orientations of sugar-phosphate chains, in which two homopyrimidine strands are parallel or antiparallel with each other, were considered in the systematic search. Several possible base-pairing models, in which there are one Watson-Crick base pair and one other base pair, were also considered. Many possible models selected by the systematic search were further refined through molecular mechanics calculation incorporating a helical boundary condition. The preferred model, which was selected on the basis of potential energy, was the one with Watson-Crick and Hoogsteen base pairs and with its two polypyrimidine chains in the antiparallel orientation. The model can explain the experimental observation that poly(L-dA) forms a stable triple helix with poly(uridylic acid) (U) but not with poly(deoxythymidylic acid) (dT).

DNA↗
PubMed2005

A case study in comparing therapies involving informative drop-out, non-ignorable non-compliance and repeated measurements.

Virtually no comparisons of different psychotherapies with long follow-up times have been carried out until now. The Helsinki Psychotherapy Study is a randomized clinical trial, where patients were monitored for 12 months after the onset of study treatments, of which each lasted approximately 6 months. The patients' psychiatric status was measured at five pre-determined time points during the follow-up period. In general, the analyses of trials are complicated in cases where compliance with the given treatment is incomplete or the drop-out from the follow-up is non-ignorable. In the present study, the quality of the treatment deviated from the protocol for some patients and some patients took auxiliary treatments which had similar effects to the study treatment during the study treatment or follow-up period. This might have resulted in standard intention-to-treat analyses providing excessively conservative or liberal conclusions. Non-compliance may have been non-ignorable in some cases, so subject-specific latent factors may have influenced the outcome both directly and indirectly via compliance behaviour. The most and least healthy patients are the most likely to dropout from the follow-up a priori, so the missing data process is informative. The missing data can partly be augmented with surrogate information collected during interviews with patients who dropped out. A Bayesian hierarchical as-treated model, which uses random-effects-based selection models to account for non-ignorable missing data and non-compliance, was compared with different mixed effects models.

Adult↗
PubMed2004

Using wildlife as receptor species: a landscape approach to ecological risk assessment.

To assist risk assessors at the Department of Energy's Savannah River Site (SRS), a Geographic Information System (GIS) application was developed to provide relevant information about specific receptor species of resident wildlife that can be used for ecological risk assessment. Information was obtained from an extensive literature review of publications and reports on vertebrate- and contaminant-related research since 1954 and linked to a GIS. Although this GIS is a useful tool for risk assessors because the data quality is high, it does not describe the species' site-wide spatial distribution or life history, which may be crucial when developing a risk assessment. Specific receptor species on the SRS were modeled to provide an estimate of an overall distribution (probability of being in an area). Each model is a stand-alone tool consisting of algorithms independent of the GIS data layers to which it is applied and therefore is dynamic and will respond to changes such as habitat disturbances and natural succession. This paper describes this modeling process and demonstrates how these resource selection models can then be used to produce spatially explicit exposure estimates. This approach is a template for other large federal facilities to establish a framework for site-specific risk assessments that use wildlife species as endpoints.

Animals↗
PubMed2002

Frequency and correlates of coronary stent thrombosis in the modern era: analysis of a single center registry.

OBJECTIVES: The study examined the frequency, correlates, and outcome of patients with stent thrombosis within 30 days of stent placement. BACKGROUND: Patients in trials evaluating stents or dual antiplatelet therapy to prevent coronary stent thrombosis have generally had narrow inclusion criteria; the extent to which stent thrombosis rates in such trials represent current practice, particularly with the availability of newer stents, is unclear. METHODS: We performed a retrospective analysis of the Mayo Clinic Percutaneous Coronary Intervention database and identified all patients who received at least one coronary stent and dual antiplatelet therapy (aspirin and ticlopidine or clopidogrel for two to four weeks). RESULTS: Four thousand five hundred nine patients underwent successful coronary stent implantation and were treated with dual antiplatelet therapy between July 1, 1994, and April 30, 2000. Stent thrombosis occurred in 23 patients (0.51%; 95% confidence interval 0.32%, 0.76%) within 30 days of stent placement. Multivariate analysis using bootstrap model selection to avoid over-fitting the model indicated that only the number of stents placed was an independent correlate of stent thrombosis (odds ratio 1.80, p < 0.001). The frequency of death and frequency of nonfatal myocardial infarction (MI) among the 23 patients with stent thrombosis were 48% and 39%, respectively. CONCLUSIONS: Stent thrombosis is even more rare in the current era than in earlier trials. Number of stents placed was an independent correlate of stent thrombosis. Most patients who suffer stent thrombosis either die or suffer MI.

Aged↗
PubMed1999

An anion-selective analogue of the channel-forming peptide alamethicin.

The peptide alamethicin self-assembles to form helix bundle ion channels in membranes. Previous macroscopic measurements have shown that these channels are mildly cation-selective. Models indicate that a source of cation selectivity is a zone of partial negative charge toward the C-terminal end of the peptide. We synthesized an alamethicin derivative with a lysine in this zone (replacing the glutamine at position 18 in the sequence). Microscopic (single-channel) measurements demonstrate that dimeric alamethicin-lysine18 (alm-K18) forms mildly anion-selective channels under conditions where channels formed by the parent peptide are cation-selective. Long-range electrostatic interactions can explain the inversion of ion selectivity and the conductance properties of alamethicin channels.

Alamethicin↗
PubMed2000

Derivation of pharmacophore and CoMFA models for leukotriene D(4) receptor antagonists of the quinolinyl(bridged)aryl series.

The present work focuses on the study of the three-dimensional (3D) structural requirements for the leukotriene D(4) (LTD(4)) antagonistic activity of compounds having the basic quinolinyl(bridged)aryl framework. An approach combining pharmacophore mapping, molecule alignment, and CoMFA models was used to derive a hypothesis for a series of LTD(4) antagonists having the basic diaryl-bridged framework. In this compound series, the produced pharmacophore hypotheses have shown to yield molecule alignments suitable to derive valuable CoMFA models. Model selection focused on (1) obtention of coherent modeling results, (2) consistency with the available SAR data, and (3) ability to predict the activity of an independent set of congeneric molecules. This approach resulted in a combined pharmacophore and CoMFA model that can generally represent the antagonistic activity within a log unit of the measured value for compounds of the series. The resulting pharmacophore (model C) consists of an acidic or negative ionizable function (AC), a hydrogen-bond acceptor (HBA), and three hydrophobic regions (HY) and produces chemically meaningful alignments with the most active compounds of the series mapping the pharmacophore in a extended energetically favorable conformation.

Combinatorial Chemistry Techniques↗
PubMed2001

A confirmatory factor analysis on the DSM-IV ADHD and ODD symptoms: what is the best model for the organization of these symptoms?

Confirmatory factor analysis (CFA) was used to evaluate five different models for the organization of the DSM-IV ADHD and oppositional defiant disorder (ODD) symptoms (Model 1: a single factor model; Model 2: an ADHD and ODD two factor model; Model 3a: an inattention (INA), hyperactivity/impulsivity (HYP/IMP), and ODD three factor model; Model 3b: an INA, HYP/IMP, and ODD three factor model where the three IMP symptoms cross-load on the ODD factor; Model 4: an INA, HYP, IMP, and ODD four factor model). To evaluate these models, maternal ratings of ADHD and ODD symptoms were obtained at outpatient pediatric clinics on 742 children not in treatment and 91 children in treatment for ADHD. Model 3a resulted in a good fit as well as a significantly better fit than Model 2. Model 3a was also equivalent across treatment status, gender, and age groupings for the most part. Though Models 3b and 4 provided a statistically better fit than Model 3a, the improvement in fit was small and other model selection criteria argued against these more complex models.

Attention Deficit Disorder with Hyperactivity↗
PubMed2003

A Bayesian molecular interaction library.

We describe a library of molecular fragments designed to model and predict non-bonded interactions between atoms. We apply the Bayesian approach, whereby prior knowledge and uncertainty of the mathematical model are incorporated into the estimated model and its parameters. The molecular interaction data are strengthened by narrowing the atom classification to 14 atom types, focusing on independent molecular contacts that lie within a short cutoff distance, and symmetrizing the interaction data for the molecular fragments. Furthermore, the location of atoms in contact with a molecular fragment are modeled by Gaussian mixture densities whose maximum a posteriori estimates are obtained by applying a version of the expectation-maximization algorithm that incorporates hyperparameters for the components of the Gaussian mixtures. A routine is introduced providing the hyperparameters and the initial values of the parameters of the Gaussian mixture densities. A model selection criterion, based on the concept of a 'minimum message length' is used to automatically select the optimal complexity of a mixture model and the most suitable orientation of a reference frame for a fragment in a coordinate system. The type of atom interacting with a molecular fragment is predicted by values of the posterior probability function and the accuracy of these predictions is evaluated by comparing the predicted atom type with the actual atom type seen in crystal structures. The fact that an atom will simultaneously interact with several molecular fragments forming a cohesive network of interactions is exploited by introducing two strategies that combine the predictions of atom types given by multiple fragments. The accuracy of these combined predictions is compared with those based on an individual fragment. Exhaustive validation analyses and qualitative examples (e.g., the ligand-binding domain of glutamate receptors) demonstrate that these improvements lead to effective modeling and prediction of molecular interactions.

Algorithms↗