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.

SourceResearch connectionFrom yearThrough year

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

At least 55 records · Page 3Linked to original sources

PubMed1972

Maintenance of genetic variation with a frequency-dependent selection model as compared to the overdominant model.

A frequency-dependent selection model proposed by Huang, Singh and Kojima (1971) was found to be more effective at maintaining genetic variation in a finite population than the overdominant model. The fourth moment parameter of the distribution of unfixed states showed that there was a more platykurtic distribution for the frequency-dependent model. This agreed well with the expected gene frequency change found for an infinite population.

Analysis of Variance↗

PubMed2004

Model-free analysis of protein dynamics: assessment of accuracy and model selection protocols based on molecular dynamics simulation.

The popular model-free approach to analyze NMR relaxation measurements has been examined using artificial amide (15)N relaxation data sets generated from a 10 nanosecond molecular dynamics trajectory of a dihydrofolate reductase ternary complex in explicit water. With access to a detailed picture of the underlying internal motions, the efficacy of model-free analysis and impact of model selection protocols on the interpretation of NMR data can be studied. In the limit of uncorrelated global tumbling and internal motions, fitting the relaxation data to the model-free models can recover a significant amount of quantitative information on the internal dynamics. Despite a slight overestimation, the generalized order parameter is quite accurately determined. However, the model-free analysis appears to be insensitive to the presence of nanosecond time scale motions with relatively small magnitude. For such cases, the effective correlation time can be significantly underestimated. As a result, proteins appear to be more rigid than they really are. The model selection protocols have a major impact on the information one can reliably obtain. The commonly employed protocol based on step-up hypothesis testing has severe drawbacks of oversimplification and underfitting. The consequences are that the order parameter is more severely overestimated and the correlation time more severely underestimated. Instead, model selection based on Bayesian Information Criteria (BIC), recently introduced to the model-free analysis by d'Auvergne and Gooley (2003), provides a better balance between bias and variance. More appropriate models can be selected, leading to improved estimate of both the order parameter and correlation time. In addition, the computational cost is significantly reduced and subjective parameters such as the significance level are unnecessary.

Anisotropy↗

PubMed1997

Discriminant analysis using the unweighted sum of binary variables: a comparison of model selection methods.

Many clinical decision-making rules are equivalent to linear discriminant functions that involve the unweighted sum of binary variables (SBV). We briefly consider the geometry of this restriction and then propose a number of methods for forward stepwise selection of SBV models. Using a simulation study, we compare the performance of these methods under a wide range of plausible conditions and show that no single method is uniformly superior for selecting models of a fixed size. Factors of general importance in relative method performance are the ratio of sample size to the number of candidate variables and the class-conditional moment structure of the data. We conclude by offering some practical strategies for SBV model construction.

Computer Simulation↗

PubMed2005

Bayesian model selection and averaging in additive and proportional hazards models.

Although Cox proportional hazards regression is the default analysis for time to event data, there is typically uncertainty about whether the effects of a predictor are more appropriately characterized by a multiplicative or additive model. To accommodate this uncertainty, we place a model selection prior on the coefficients in an additive-multiplicative hazards model. This prior assigns positive probability, not only to the model that has both additive and multiplicative effects for each predictor, but also to sub-models corresponding to no association, to only additive effects, and to only proportional effects. The additive component of the model is constrained to ensure non-negative hazards, a condition often violated by current methods. After augmenting the data with Poisson latent variables, the prior is conditionally conjugate, and posterior computation can proceed via an efficient Gibbs sampling algorithm. Simulation study results are presented, and the methodology is illustrated using data from the Framingham heart study.

Age of Onset↗

PubMed1992

A worst-case optimal parameter selection model of cancer chemotherapy.

An optimal parameter selection model of cancer chemotherapy in which two system parameters are unknown is formulated as a worst-case optimal parameter selection model. The model assumes that the unknown parameters lie within a known set. The system constraints must be satisfied over this entire set, and the objective function minimized in the worst case. The continuous dependence of the objective function and the system constraints upon the unknown parameters can be removed, making a numerical solution tractable. For the data considered it is proven that a cure is impossible no matter what the values of the unknown parameters in the parameter set. The optimal policy is shown to be relatively low dose intensity for the majority of the treatment, with the remaining drug delivered towards the end of the treatment interval.

Humans↗

PubMed2001

How carcinogens (or telomere dysfunction) induce genetic instability: associated-selection model.

Carcinogens induce carcinogen-specific genetic instability (defects in DNA repair). According to the 'direct-selection' model, defects in DNA repair per se provide an immediate growth advantage. According to the 'associated-selection' model, carcinogens merely select for cells with adaptive mutations. Like any mutations, adaptive mutations occur predominantly in genetically unstable cells. The 'associated-selection' model predicts that carcinogen-driven selection minimizes cytotoxic but maximizes mutagenic effects of carcinogens. A purely mutagenic (neither cytotoxic, nor cytostatic) environment will favor effective DNA repair, whereas any growth-limiting conditions (telomerase deficiency, anticancer drugs) will select for genetically unstable cells. Genetic instability is a postmark of selective pressure rather than a hallmark of cancer per se. Once selected, genetic instability facilitates the development of resistance to any other growth-limiting conditions. As an example, a putative link between prior exposure to carcinogens and the ability to develop a telomerase-independent growth is discussed.

Carcinogens↗

PubMed2001

Model selection in non-nested hidden Markov models for ion channel gating.

An important task in the application of Markov models to the analysis of ion channel data is the determination of the correct gating scheme of the ion channel under investigation. Some prior knowledge from other experiments can reduce significantly the number of possible models. If these models are standard statistical procedures nested like likelihood ratio testing, provide reliable selection methods. In the case of non-nested models, information criteria like AIC, BIC, etc., are used. However, it is not known if any of these criteria provide a reliable selection method and which is the best one in the context of ion channel gating. We provide an alternative approach to model selection in the case of non-nested models with an equal number of open and closed states. The models to choose from are embedded in a properly defined general model. Therefore, we circumvent the problems of model selection in the non-nested case and can apply model selection procedures for nested models.

Animals↗

PubMed2003

Comparison of model selection for regression.

We discuss empirical comparison of analytical methods for model selection. Currently, there is no consensus on the best method for finite-sample estimation problems, even for the simple case of linear estimators. This article presents empirical comparisons between classical statistical methods - Akaike information criterion (AIC) and Bayesian information criterion (BIC) - and the structural risk minimization (SRM) method, based on Vapnik-Chervonenkis (VC) theory, for regression problems. Our study is motivated by empirical comparisons in Hastie, Tibshirani, and Friedman (2001), which claims that the SRM method performs poorly for model selection and suggests that AIC yields superior predictive performance. Hence, we present empirical comparisons for various data sets and different types of estimators (linear, subset selection, and k-nearest neighbor regression). Our results demonstrate the practical advantages of VC-based model selection; it consistently outperforms AIC for all data sets. In our study, SRM and BIC methods show similar predictive performance. This discrepancy (between empirical results obtained using the same data) is caused by methodological drawbacks in Hastie et al. (2001), especially in their loose interpretation and application of SRM method. Hence, we discuss methodological issues important for meaningful comparisons and practical application of SRM method. We also point out the importance of accurate estimation of model complexity (VC-dimension) for empirical comparisons and propose a new practical estimate of model complexity for k-nearest neighbors regression.

Bayes Theorem↗

PubMed1997

A Bayesian Approach to Model Selection in Hierarchical Mixtures-of-Experts Architectures.

There does not exist a statistical model that shows good performance on all tasks. Consequently, the model selection problem is unavoidable; investigators must decide which model is best at summarizing the data for each task of interest. This article presents an approach to the model selection problem in hierarchical mixtures-of-experts architectures. These architectures combine aspects of generalized linear models with those of finite mixture models in order to perform tasks via a recursive "divide-and-conquer" strategy. Markov chain Monte Carlo methodology is used to estimate the distribution of the architectures' parameters. One part of our approach to model selection attempts to estimate the worth of each component of an architecture so that relatively unused components can be pruned from the architecture's structure. A second part of this approach uses a Bayesian hypothesis testing procedure in order to differentiate inputs that carry useful information from nuisance inputs. Simulation results suggest that the approach presented here adheres to the dictum of Occam's razor; simple architectures that are adequate for summarizing the data are favored over more complex structures. Copyright 1997 Elsevier Science Ltd. All Rights Reserved.

Journal Article↗

PubMed2002

On the problem in model selection of neural network regression in overrealizable scenario.

In considering a statistical model selection of neural networks and radial basis functions under an overrealizable case, the problem of unidentifiability emerges. Because the model selection criterion is an unbiased estimator of the generalization error based on the training error, this article analyzes the expected training error and the expected generalization error of neural networks and radial basis functions in overrealizable cases and clarifies the difference from regular models, for which identifiability holds. As a special case of an overrealizable scenario, we assumed a gaussian noise sequence as training data. In the least-squares estimation under this assumption, we first formulated the problem, in which the calculation of the expected errors of unidentifiable networks is reduced to the calculation of the expectation of the supremum of the chi2 process. Under this formulation, we gave an upper bound of the expected training error and a lower bound of the expected generalization error, where the generalization is measured at a set of training inputs. Furthermore, we gave stochastic bounds on the training error and the generalization error. The obtained upper bound of the expected training error is smaller than in regular models, and the lower bound of the expected generalization error is larger than in regular models. The result tells us that the degree of overfitting in neural networks and radial basis functions is higher than in regular models. Correspondingly, it also tells us that the generalization capability is worse than in the case of regular models. The article may be enough to show a difference between neural networks and regular models in the context of the least-squares estimation in a simple situation. This is a first step in constructing a model selection criterion in an overrealizable case. Further important problems in this direction are also included in this article.

Journal Article↗

PubMed1991

Tests and model selection for the general growth curve model.

The model considered here is a generalized multivariate analysis of variance model useful especially for many types of growth curve problems including biological growth and technology substitution. It is defined as Yp x N = Xp x m tau m x r Ar x N + epsilon p x N, where tau is unknown, and X and A are known design matrices of ranks m less than p and r less than N, respectively. Furthermore, the columns of epsilon are independent p-variate normal with mean vector 0 and common covariance matrix sigma. In general, p is the number of time (or spatial) points observed on each of the N cases, (m - 1) is the degree of polynomial in time, and r is the number of groups. The main focus of this paper is the selection of models for the general growth curve model with regard to the covariance matrix sigma. Likelihood ratio tests and selection procedures based on sample reuse and predictions are proposed. Special emphasis is on the serial covariance structure for sigma, which has been shown to be quite important in the prediction of biological data and technology substitution data. One-population and K-population problems are considered. Some of the results are illustrated with two sets of biological data.

Animals↗

PubMed1995

Model selection for extended quasi-likelihood models in small samples.

We develop a small sample criterion (AICc) for the selection of extended quasi-likelihood models. In contrast to the Akaike information criterion (AIC). AICc provides a more nearly unbiased estimator for the expected Kullback-Leibler information. Consequently, it often selects better models than AIC in small samples. For the logistic regression model, Monte Carlo results show that AICc outperforms AIC, Pregibon's (1979, Data Analytic Methods for Generalized Linear Models. Ph.D. thesis. University of Toronto) Cp*, and the Cp selection criteria of Hosmer et al. (1989, Biometrics 45, 1265-1270). Two examples are presented.

Age Factors↗

PubMed2001

Best harmony, unified RPCL and automated model selection for unsupervised and supervised learning on Gaussian mixtures, three-layer nets and ME-RBF-SVM models.

After introducing the fundamentals of BYY system and harmony learning, which has been developed in past several years as a unified statistical framework for parameter learning, regularization and model selection, we systematically discuss this BYY harmony learning on systems with discrete inner-representations. First, we shown that one special case leads to unsupervised learning on Gaussian mixture. We show how harmony learning not only leads us to the EM algorithm for maximum likelihood (ML) learning and the corresponding extended KMEAN algorithms for Mahalanobis clustering with criteria for selecting the number of Gaussians or clusters, but also provides us two new regularization techniques and a unified scheme that includes the previous rival penalized competitive learning (RPCL) as well as its various variants and extensions that performs model selection automatically during parameter learning. Moreover, as a by-product, we also get a new approach for determining a set of 'supporting vectors' for Parzen window density estimation. Second, we shown that other special cases lead to three typical supervised learning models with several new results. On three layer net, we get (i) a new regularized ML learning, (ii) a new criterion for selecting the number of hidden units, and (iii) a family of EM-like algorithms that combines harmony learning with new techniques of regularization. On the original and alternative models of mixture-of-expert (ME) as well as radial basis function (RBF) nets, we get not only a new type of criteria for selecting the number of experts or basis functions but also a new type of the EM-like algorithms that combines regularization techniques and RPCL learning for parameter learning with either least complexity nature on the original ME model or automated model selection on the alternative ME model and RBF nets. Moreover, all the results for the alternative ME model are also applied to other two popular nonparametric statistical approaches, namely kernel regression and supporting vector machine. Particularly, not only we get an easily implemented approach for determining the smoothing parameter in kernel regression, but also we get an alternative approach for deciding the set of supporting vectors in supporting vector machine.

Algorithms↗

PubMed2004

Model selection in ecology and evolution.

Recently, researchers in several areas of ecology and evolution have begun to change the way in which they analyze data and make biological inferences. Rather than the traditional null hypothesis testing approach, they have adopted an approach called model selection, in which several competing hypotheses are simultaneously confronted with data. Model selection can be used to identify a single best model, thus lending support to one particular hypothesis, or it can be used to make inferences based on weighted support from a complete set of competing models. Model selection is widely accepted and well developed in certain fields, most notably in molecular systematics and mark-recapture analysis. However, it is now gaining support in several other areas, from molecular evolution to landscape ecology. Here, we outline the steps of model selection and highlight several ways that it is now being implemented. By adopting this approach, researchers in ecology and evolution will find a valuable alternative to traditional null hypothesis testing, especially when more than one hypothesis is plausible.

Journal Article↗

PubMed1998

The application of sample selection models to outcomes research: the case of evaluating the effects of antidepressant therapy on resource utilization.

Non-randomized studies of treatment effects have come under criticism because of their failure to control for potential biases introduced by unobserved variables correlated with treatment selection and outcomes. This paper describes the basic concepts of sample selection models--a technique used widely in the economics evaluation literature for nearly two decades--and discusses the potential role of these models in outcomes research. In addition, it presents a case study of the application of the sample selection modelling approach to evaluation of the effects of antidepressant therapies on medical expenditures for physician services. This case study presents empirical comparisons of alternative model specifications and discusses practical issues in evaluation of sample selection models. We demonstrate that, in this particular case, sample selection models yield very different conclusions regarding treatment effects than traditional ordinary least squares regression.

Antidepressive Agents↗

PubMed2003

Flexibility versus generalizability in model selection.

Which quantitative method should be used to choose among competing mathematical models of cognition? Massaro, Cohen, Campbell, and Rodriguez (2001) favor root mean squared deviation (RMSD), choosing the model that provides the best fit to the data. Their simulation results appear to legitimize its use for comparing two models of information integration because it performed just as well as Bayesian model selection (BMS), which had previously been shown by Myung and Pitt (1997) to be a superior alternative selection method because it considers a model's complexity in addition to its fit. In the present study, after contrasting the theoretical approaches to model selection espoused by Massaro et al. and Myung and Pitt, we discuss the cause of the inconsistencies by expanding on the simulations of Massaro et al. Findings demonstrate that the results from model recovery simulations can be misleading if they are not interpreted relative to the data on which they were evaluated, and that BMS is a more robust selection method.

Bayes Theorem↗

PubMed1988

A general asymptotic property of two-locus selection models.

It is shown that any two-locus, two-allele model of selection with constant fitnesses has at least one polymorphic equilibrium for which the linkage association measure, D, is arbitrarily close to zero for large enough recombination, R. As R----+/- infinity, D----0 in such a way that the product l = RD----a non-zero finite constant. There may be 1, 3, or 5 distinct asymptotic equilibria, depending upon fitness parameters.

Alleles↗

PubMed2006

Why kin and group selection models may not be enough to explain human other-regarding behaviour.

Models of kin or group selection usually feature only one possible fitness transfer. The phenotypes are either to make this transfer or not to make it and for any given fitness transfer, Hamilton's rule predicts which of the two phenotypes will spread. In this article we allow for the possibility that different individuals or different generations face similar, but not necessarily identical possibilities for fitness transfers. In this setting, phenotypes are preference relations, which concisely specify behaviour for a range of possible fitness transfers (rather than being a specification for only one particular situation an animal or human can be in). For this more general set-up, we find that only preference relations that are linear in fitnesses can be explained using models of kin selection and that the same applies to a large class of group selection models. This provides a new implication of hierarchical selection models that could in principle falsify them, even if relatedness--or a parameter for assortativeness--is unknown. The empirical evidence for humans suggests that hierarchical selection models alone are not enough to explain their other-regarding or altruistic behaviour.

Altruism↗