In this paper we propose a crossover operator for evolutionary algorithms with real values that is based on the statistical theory of population distributions. The operator is based on the theoretical distribution of the values of the genes of the be
st individuals in the population. The proposed operator takes into account the localization and dispersion features of the best individuals of the population with the objective that these features would be inherited by the offspring. Our aim is the optimization of the balance between exploration and exploitation in the search process. In order to test the efficiency and robustness of this crossover, we have used a set of functions to be optimized with regard to different criteria, such as, multimodality, separability, regularity and epistasis. With this set of functions we can extract conclusions in function of the problem at hand. We analyze the results using ANOVA and multiple comparison statistical tests. As an example of how our crossover can be used to solve artificial intelligence problems, we have applied the proposed model to the problem of obtaining the weight of each network in a ensemble of neural networks. The results obtained are above the performance of standard methods.
In this work we show that for a quasi-2D system of size $Omega$ and thickness $t$ the resistance goes as $(2rho/pi t)ln(Omega/W)$, diverging logarithmically with the size. Measurements in highly oriented pyrolytic graphite (HOPG) as well as numerical
simulations confirm this relation. Furthermore, we present an experimental method that allows us to obtain the carriers mean free path $l(T)$, the Fermi wavelength $lambda(T)$ and the mobility $mu(T)$ directly from experiments without adjustable parameters. Measuring the electrical resistance through microfabricated constrictions in HOPG and observing the transition from ohmic to ballistic regime we obtain that $0.2 mu$m $lesssim l lesssim 10 mu$m, $0.1 mu$m $lesssim lambda lesssim 2 mu$m and a mobility $5 times 10^4$ cm$^2$/Vs $ lesssim mu lesssim 4 times 10^7$ cm$^2$/Vs when the temperature decreases from 270K to 3K. A comparison of these results with those from literature indicates that conventional, multiband Boltzmann-Drude approaches are inadequate for oriented graphite. The upper value obtained for the mobility is much larger than the mobility graphene samples of micrometer size can have.
High resolution magnetoresistance data in highly oriented pyrolytic graphite thin samples manifest non-homogenous superconductivity with critical temperature $T_c sim 25 $K. These data exhibit: i) hysteretic loops of resistance versus magnetic field
similar to Josephson-coupled grains, ii) quantum Andreevs resonances and iii) absence of the Schubnikov-de Haas oscillations. The results indicate that graphite is a system with non-percolative superconducting domains immersed in a semiconducting-like matrix. As possible origin of the superconductivity in graphite we discuss interior-gap superconductivity when two very different electronic masses are present.
We present a study of the magnetoresistance of highly oriented pyrolytic graphite (HOPG) as a function of the sample size. Our results show unequivocally that the magnetoresistance reduces with the sample size even for samples of hundreds of micromet
ers size. This sample size effect is due the large mean free path and Fermi wavelength of carriers in graphite and may explain the observed practically absence of magnetoresistance in micrometer confined small graphene samples where quantum effects should be at hand. These were not taken into account in the literature yet and ask for a revision of experimental and theoretical work on graphite.
