Selection (genetic algorithm)

About: Selection (genetic algorithm) is a research topic. Over the lifetime, 72443 publications have been published within this topic receiving 1327417 citations.

Bayesian Model Selection in Social Research

Abstract: It is argued that P-values and the tests based upon them give unsatisfactory results, especially in large samples. It is shown that, in regression, when there are many candidate independent variables, standard variable selection procedures can give very misleading results. Also, by selecting a single model, they ignore model uncertainty and so underestimate the uncertainty about quantities of interest. The Bayesian approach to hypothesis testing, model selection, and accounting for model uncertainty is presented. Implementing this is straightforward through the use of the simple and accurate BIC approximation, and it can be done using the output from standard software. Specific results are presented for most of the types of model commonly used in sociology. It is shown that this approach overcomes the difficulties with P-values and standard model selection procedures based on them. It also allows easy comparison of nonnested models, and permits the quantification of the evidence for a null hypothesis of interest, such as a convergence theory or a hypothesis about societal norms.

The measurement of selection on correlated characters

Abstract: Natural selection acts on phenotypes, regardless of their genetic basis, and produces immediate phenotypic effects within a generation that can be measured without recourse to principles of heredity or evolution. In contrast, evolutionary response to selection, the genetic change that occurs from one generation to the next, does depend on genetic variation. Animal and plant breeders routinely distinguish phenotypic selection from evolutionary response to selection (Mayo, 1980; Falconer, 1981). Upon making this critical distinction, emphasized by Haldane (1954), precise methods can be formulated for the measurement of phenotypic natural selection. Correlations between characters seriously complicate the measurement of phenotypic selection, because selection on a particular trait produces not only a direct effect on the distribution of that trait in a population, but also produces indirect effects on the distribution of correlated characters. The problem of character correlations has been largely ignored in current methods for measuring natural selection on quantitative traits. Selection has usually been treated as if it acted only on single characters (e.g., Haldane, 1954; Van Valen, 1965a; O'Donald, 1968, 1970; reviewed by Johnson, 1976 Ch. 7). This is obviously a tremendous oversimplification, since natural selection acts on many characters simultaneously and phenotypic correlations between traits are ubiquitous. In an important but neglected paper, Pearson (1903) showed that multivariate statistics could be used to disentangle the direct and indirect effects of selection to determine which traits in a correlated ensemble are the focus of direct selection. Here we extend and generalize Pearson's major results. The purpose of this paper is to derive measures of directional and stabilizing (or disruptive) selection on each of a set of phenotypically correlated characters. The analysis is retrospective, based on observed changes in the multivariate distribution of characters within a generation, not on the evolutionary response to selection. Nevertheless, the measures we propose have a close connection with equations for evolutionary change. Many other commonly used measures of the intensity of selection (such as selective mortality, change in mean fitness, variance in fitness, or estimates of particular forms of fitness functions) have little predictive value in relation to evolutionary change in quantitative traits. To demonstrate the utility of our approach, we analyze selection on four morphological characters in a population of pentatomid bugs during a brief period of high mortality. We also summarize a multivariate selection analysis on nine morphological characters of house sparrows caught in a severe winter storm, using the classic data of Bumpus (1899). Direct observations and measurements of natural selection serve to clarify one of the major factors of evolution. Critiques of the "adaptationist program" (Lewontin, 1978; Gould and Lewontin, 1979) stress that adaptation and selection are often invoked without strong supporting evidence. We suggest quantitative measurements of selection as the best alternative to the fabrication of adaptive scenarios. Our optimism that measurement can replace rhetorical claims for adaptation and selection is founded in the growing success of field workers in their efforts to measure major components of fitness in natural populations (e.g., Thornhill, 1976; Howard, 1979; Downhower and Brown, 1980; Boag and Grant, 1981; Clutton-Brock et

PartitionFinder: Combined Selection of Partitioning Schemes and Substitution Models for Phylogenetic Analyses

Abstract: In phylogenetic analyses of molecular sequence data, partitioning involves estimating independent models of molecular evolution for different sets of sites in a sequence alignment. Choosing an appropriate partitioning scheme is an important step in most analyses because it can affect the accuracy of phylogenetic reconstruction. Despite this, partitioning schemes are often chosen without explicit statistical justification. Here, we describe two new objective methods for the combined selection of best-fit partitioning schemes and nucleotide substitution models. These methods allow millions of partitioning schemes to be compared in realistic time frames and so permit the objective selection of partitioning schemes even for large multilocus DNA data sets. We demonstrate that these methods significantly outperform previous approaches, including both the ad hoc selection of partitioning schemes (e.g., partitioning by gene or codon position) and a recently proposed hierarchical clustering method. We have implemented these methods in an open-source program, PartitionFinder. This program allows users to select partitioning schemes and substitution models using a range of information-theoretic metrics (e.g., the Bayesian information criterion, akaike information criterion [AIC], and corrected AIC). We hope that PartitionFinder will encourage the objective selection of partitioning schemes and thus lead to improvements in phylogenetic analyses. PartitionFinder is written in Python and runs under Mac OSX 10.4 and above. The program, source code, and a detailed manual are freely available from www.robertlanfear.com/partitionfinder.