Search results for "Genetic algorithm"
showing 10 items of 834 documents
Improving Pattern Recognition Based Pharmacological Drug Selection Through ROC Analysis
2004
The design of new medical drugs is a very complex process in which combinatorial chemistry techniques are used. The goal consists of discriminating between molecular compounds exhibiting or not certain pharmacological activities. Different machine learning approaches have been recently applied to different drug design problems leading to competitive results in pointing at particular compounds with high probability of exhibiting activity. The present work first deeps into the natural trade-off between accuracy in the much less populated active group and false alarm rate which could lead to too many expensive laboratory tests. Preliminary results show how different classification techniques a…
Optimal Usage of Multiple Network Connections
2008
In the future mobile networks, a mobile terminal is able to select the best suitable network for each data transmission. The selection of a network connection to be used has been under a lot of study. In this paper, we consider a more extensive case in which we do not select a network connection but use several network connections simultaneously to transfer data. When data is transferred using multiple network connections, a network connection has to be selected for each component of the data. We have modelled this problem as a multiobjective optimization problem and developed a heuristic to solve the problem fast in a static network environment. In this paper, we discuss solving the proble…
Niche filling slows the diversification of Himalayan songbirds.
2013
In Himalayan songbirds, the speciation rate is ultimately set by ecological competition, rather than by the rate of acquisition of reproductive isolation. The beginnings of adaptive radiation and speciation have been widely studied — in Darwin's finches, sticklebacks and cichlid fish, for example — but relatively little is known about what happens next. Specifically, what is the rate-limiting step for the establishment of new species? This seven-year study of the 358 songbird species found on the Himalayan slopes suggests that it is the rates at which new niches are created and occupied that limits diversification, not the rate at which new species form through reproductive isolation. Speci…
Conditions for the spread of conspicuous warning signals: A numerical model with novel insights
2007
The initial evolution of conspicuous warning signals presents an evolutionary problem because selection against rare conspicuous signals is presumed to be strong, and new signals are rare when they first arise. Several possible solutions have been offered to solve this apparent evolutionary paradox, but disagreement persists over the plausibility of some of the proposed mechanisms. In this paper, we construct a deterministic numerical simulation model that allows us to derive the strength of selection on novel warning signals in a wide range of biologically relevant situations. We study the effects of predator psychology (learning, rate of mistaken attacks, and neophobia) on selection. We a…
SELECTING HERB-RICH FOREST NETWORKS TO PROTECT DIFFERENT MEASURES OF BIODIVERSITY
2001
Data on vascular plants of herb-rich forests in Finland were used to compare the efficiency of reserve selection methods in representing three measures of biodiversity: species richness, phylogenetic diversity, and restricted-range diversity. Comparisons of reserve selection methods were carried out both with and without consideration of the existing reserve system. Our results showed that the success of a reserve network of forests in representing different measures of biodiversity depends on the selection procedure, selection criteria, and data set used. Ad hoc selection was the worst option. A scoring procedure was generally more efficient than maximum random selection. Heuristic methods…
Venture Capitalists' Decision to Syndicate.
2006
International audience; Financial theory, access to deal flow, selection, and monitoring skills are used to explain syndication in venture capital firms in six European countries. In contrast with U.S. findings, portfolio management motives are more important for syndication than individual deal management motives. Risk sharing, portfolio diversification, and access to larger deals are more important than selection and monitoring of deals. This holds for later stage and for early stage investors. Value adding is a stronger motive for syndication for early stage investors than for later stage investors, however. Nonlead investors join syndicates for the selection and value-adding skills of t…
Selection by committee: Anonymity and gratitude
2018
Abstract What kind of candidate is selected into a job when the principal has to appoint a committee to measure the candidates’ ability and select a winner? We find that if the committee takes into account the candidate’s gratitude towards them, a candidate with less than first best ability will be selected. A relevant exception may occur if the first best is the overall best candidate. First best selection is always achieved if the committee is anonymous to the candidates. If the committee is not detached enough from the candidates then delegation fares even worse than random selection.
Does Firm Size Affect Self-selection and Learning-by-Exporting?
2010
The trade literature has long discussed the existence of some benefits attributed to exporting, among others, the improvement of firm productivity. This paper examines whether firm size plays a role in this supposedly favourable relationship between exporting and total factor productivity (TFP). To examine this, we investigate, separately for large and small firms, whether firms starting to export perform better ex ante (self-selection) than non-exporting firms and, conditional on this fact, if they are also more productive ex post (learning-by-exporting). With this purpose, we use both stochastic dominance and matching techniques. The dataset is a representative sample of Spanish manufactu…
A problem-adjusted genetic algorithm for flexibility design
2013
Many present markets for goods and services have highly volatile demand due to short life cycles and strong competition in saturated environments. Determination of capacity levels is difficult because capacities often need to be set long before demand realizes. In order to avoid capacity-demand mismatches, operations managers employ mix-flexible resources which allow them to shift excess demands to unused capacities. The Flexibility Design Problem (FDP) models the decision on the optimal configuration of a flexible (manufacturing) network. FDP is a difficult stochastic optimization problem, for which traditional exact approaches are not able to solve but the smallest instances in reasonable…
A naïve approach to speed up portfolio optimization problem using a multiobjective genetic algorithm
2012
a b s t r a c t Genetic algorithms (GAs) are appropriate when investors have the objective of obtaining mean-variance (VaR) efficient frontier as minimising VaR leads to non-convex and non-differential risk-return optimisation problems. However GAs are a time-consuming optimisation technique. In this paper, we propose to use a naive approach consisting of using samples split by quartile of risk to obtain complete efficient frontiers in a reasonable computation time. Our results show that using reduced problems which only consider a quartile of the assets allow us to explore the efficient frontier for a large range of risk values. In particular, the third quartile allows us to obtain efficie…