Combining meta-learning and search techniques to select parameters for support vector machines

Support Vector Machines (SVMs) have achieved very good performance on different learning problems. However, the success of SVMs depends on the adequate choice of the values of a number of parameters (e.g., the kernel and regularization parameters). In the current work, we propose the combination of meta-learning and search algorithms to deal with the problem of SVM parameter selection. In this combination, given a new problem to be solved, meta-learning is employed to recommend SVM parameter values based on parameter configurations that have been successfully adopted in previous similar problems. The parameter values returned by meta-learning are then used as initial search points by a search technique, which will further explore the parameter space. In this proposal, we envisioned that the initial solutions provided by meta-learning are located in good regions of the search space (i.e. they are closer to optimum solutions). Hence, the search algorithm would need to evaluate a lower number of candidate solutions when looking for an adequate solution. In this work, we investigate the combination of meta-learning with two search algorithms: Particle Swarm Optimization and Tabu Search. The implemented hybrid algorithms were used to select the values of two SVM parameters in the regression domain. These combinations were compared with the use of the search algorithms without meta-learning. The experimental results on a set of 40 regression problems showed that, on average, the proposed hybrid methods obtained lower error rates when compared to their components applied in isolation.

Efficient Forest Data Structure for Evolutionary Algorithms Applied to Network Design

The design of a network is a solution to several engineering and science problems. Several network design problems are known to be NP-hard, and population-based metaheuristics like evolutionary algorithms (EAs) have been largely investigated for such problems. Such optimization methods simultaneously generate a large number of potential solutions to investigate the search space in breadth and, consequently, to avoid local optima. Obtaining a potential solution usually involves the construction and maintenance of several spanning trees, or more generally, spanning forests. To efficiently explore the search space, special data structures have been developed to provide operations that manipulate a set of spanning trees (population). For a tree with n nodes, the most efficient data structures available in the literature require time O(n) to generate a new spanning tree that modifies an existing one and to store the new solution. We propose a new data structure, called node-depth-degree representation (NDDR), and we demonstrate that using this encoding, generating a new spanning forest requires average time O(root n). Experiments with an EA based on NDDR applied to large-scale instances of the degree-constrained minimum spanning tree problem have shown that the implementation adds small constants and lower order terms to the theoretical bound.

Investigating Smart Sampling as a population initialization method for Differential Evolution in continuous problems

Recently, researches have shown that the performance of metaheuristics can be affected by population initialization. Opposition-based Differential Evolution (ODE), Quasi-Oppositional Differential Evolution (QODE), and Uniform-Quasi-Opposition Differential Evolution (UQODE) are three state-of-the-art methods that improve the performance of the Differential Evolution algorithm based on population initialization and different search strategies. In a different approach to achieve similar results, this paper presents a technique to discover promising regions in a continuous search-space of an optimization problem. Using machine-learning techniques, the algorithm named Smart Sampling (SS) finds regions with high possibility of containing a global optimum. Next, a metaheuristic can be initialized inside each region to find that optimum. SS and DE were combined (originating the SSDE algorithm) to evaluate our approach, and experiments were conducted in the same set of benchmark functions used by ODE, QODE and UQODE authors. Results have shown that the total number of function evaluations required by DE to reach the global optimum can be significantly reduced and that the success rate improves if SS is employed first. Such results are also in consonance with results from the literature, stating the importance of an adequate starting population. Moreover, SS presents better efficacy to find initial populations of superior quality when compared to the other three algorithms that employ oppositional learning. Finally and most important, the SS performance in finding promising regions is independent of the employed metaheuristic with which SS is combined, making SS suitable to improve the performance of a large variety of optimization techniques. (C) 2012 Elsevier Inc. All rights reserved.

Transiting exoplanets from the CoRoT space mission XIX. CoRoT-23b: a dense hot Jupiter on an eccentric orbit

We report the detection of CoRoT-23b, a hot Jupiter transiting in front of its host star with a period of 3.6314 +/- 0.0001 days. This planet was discovered thanks to photometric data secured with the CoRoT satellite, combined with spectroscopic radial velocity (RV) measurements. A photometric search for possible background eclipsing binaries conducted at CFHT and OGS concluded with a very low risk of false positives. The usual techniques of combining RV and transit data simultaneously were used to derive stellar and planetary parameters. The planet has a mass of M-p = 2.8 +/- 0.3 M-Jup, a radius of R-pl = 1.05 +/- 0.13 R-Jup, a density of approximate to 3 gcm(-3). RV data also clearly reveal a nonzero eccentricity of e = 0.16 +/- 0.02. The planet orbits a mature G0 main sequence star of V = 15.5 mag, with a mass M-star = 1.14 +/- 0.08 M-circle dot, a radius R-star = 1. 61 +/- 0.18 R-circle dot and quasi-solar abundances. The age of the system is evaluated to be 7 Gyr, not far from the transition to subgiant, in agreement with the rather large stellar radius. The two features of a significant eccentricity of the orbit and of a fairly high density are fairly uncommon for a hot Jupiter. The high density is, however, consistent with a model of contraction of a planet at this mass, given the age of the system. On the other hand, at such an age, circularization is expected to be completed. In fact, we show that for this planetary mass and orbital distance, any initial eccentricity should not totally vanish after 7 Gyr, as long as the tidal quality factor Q(p) is more than a few 10(5), a value that is the lower bound of the usually expected range. Even if CoRoT-23b features a density and an eccentricity that are atypical of a hot Jupiter, it is thus not an enigmatic object.

Transiting exoplanets from the CoRoT space mission XXIII. CoRoT-21b: a doomed large Jupiter around a faint subgiant star

CoRoT-21, a F8IV star of magnitude V = 16 mag, was observed by the space telescope CoRoT during the Long Run 01 ( LRa01) in the first winter field (constellation Monoceros) from October 2007 to March 2008. Transits were discovered during the light curve processing. Radial velocity follow-up observations, however, were performed mainly by the 10-m Keck telescope in January 2010. The companion CoRoT-21b is a Jupiter-like planet of 2.26 +/- 0.33 Jupiter masses and 1.30 +/- 0.14 Jupiter radii in an circular orbit of semi-major axis 0.0417 +/- 0.0011 AU and an orbital period of 2.72474 +/- 0.00014 days. The planetary bulk density is ( 1.36 +/- 0.48) x 10(3) kg m(-3), very similar to the bulk density of Jupiter, and follows an M-1/3 - R relation like Jupiter. The F8IV star is a sub-giant star of 1.29 +/- 0.09 solar masses and 1.95 +/- 0.2 solar radii. The star and the planet exchange extreme tidal forces that will lead to orbital decay and extreme spin-up of the stellar rotation within 800 Myr if the stellar dissipation is Q(*)/k2(*) <= 107.

Rates of description of Phytoseiidae mite species (Acari: Mesostigmata): space, time and body size variations

This study aims to analyse the degree of completeness of world inventory of the mite family Phytoseiidae and the factors that might determine the process of species description. The world data set includes 2,122 valid species described from 1839 to 2010. Species accumulation curves were analysed. The effect of localisation (latitude ranges) and body size on the species description patterns over space and time was assessed. A low proportion of species seems remain to be described, but this trend could be explained by a critical reduction in the number of specialists dedicated to the study of those mites. In addition, this trend refers to the areas where phytoseiids have been well studied around the world, and it may change considerably if the study of these mites would be intensified in some areas. The number of newly described species is lower near the tropics, and their body size is also smaller. Differences in body size were noted between the three sub-families of Phytoseiidae, the highest mean body lengths of adult females being observed for Amblyseiinae, the most diverse family. In the future, collections would have certainly to take into consideration such conclusions for instance in using more adequate optical equipment especially for field collections. The decrease in the number of phytoseiid mite described was confirmed and the factors that could explain such a trend are discussed. Information for improving further inventories is provided and discussed, especially in relation to sampling localization and study methods.