Algoritmo Q-learning como estratégia de exploração e/ou explotação para metaheurísticas GRASP e algoritmo genético

Techniques of optimization known as metaheuristics have achieved success in the resolution of many problems classified as NP-Hard. These methods use non deterministic approaches that reach very good solutions which, however, don t guarantee the determination of the global optimum. Beyond the inherent difficulties related to the complexity that characterizes the optimization problems, the metaheuristics still face the dilemma of xploration/exploitation, which consists of choosing between a greedy search and a wider exploration of the solution space. A way to guide such algorithms during the searching of better solutions is supplying them with more knowledge of the problem through the use of a intelligent agent, able to recognize promising regions and also identify when they should diversify the direction of the search. This way, this work proposes the use of Reinforcement Learning technique - Q-learning Algorithm - as exploration/exploitation strategy for the metaheuristics GRASP (Greedy Randomized Adaptive Search Procedure) and Genetic Algorithm. The GRASP metaheuristic uses Q-learning instead of the traditional greedy-random algorithm in the construction phase. This replacement has the purpose of improving the quality of the initial solutions that are used in the local search phase of the GRASP, and also provides for the metaheuristic an adaptive memory mechanism that allows the reuse of good previous decisions and also avoids the repetition of bad decisions. In the Genetic Algorithm, the Q-learning algorithm was used to generate an initial population of high fitness, and after a determined number of generations, where the rate of diversity of the population is less than a certain limit L, it also was applied to supply one of the parents to be used in the genetic crossover operator. Another significant change in the hybrid genetic algorithm is the proposal of a mutually interactive cooperation process between the genetic operators and the Q-learning algorithm. In this interactive/cooperative process, the Q-learning algorithm receives an additional update in the matrix of Q-values based on the current best solution of the Genetic Algorithm. The computational experiments presented in this thesis compares the results obtained with the implementation of traditional versions of GRASP metaheuristic and Genetic Algorithm, with those obtained using the proposed hybrid methods. Both algorithms had been applied successfully to the symmetrical Traveling Salesman Problem, which was modeled as a Markov decision process

A expressão do condicionado contrafactual em construções se p, então q no português brasileiro

Conselho Nacional de Desenvolvimento Científico e Tecnológico

Coenzyme Q(10) and its effects in the treatment of neurodegenerative diseases

According to clinical and pre-clinical studies, oxidative stress and its consequences may be the cause or, at least, a contributing factor, to a large number of neurodegenerative diseases. These diseases include common and debilitating disorders, characterized by progressive and irreversible loss of neurons in specific regions of the brain. The most common neurodegenerative diseases are Parkinson's disease, Huntington's disease, Alzheimer's disease and amyotrophic lateral sclerosis. Coenzyme Q(10) (CoQ(10)) has been extensively studied since its discovery in 1957. It is a component of the electron transportation chain and participates in aerobic cellular respiration, generating energy in the form of adenosine triphosphate (ATP). The property of CoQ(10) to act as an antioxidant or a pro-oxidant, suggests that it also plays an important role in the modulation of redox cellular status under physiological and pathological conditions, also performing a role in the ageing process. In several animal models of neurodegenerative diseases, CoQ(10) has shown beneficial effects in reducing disease progression. However, further studies are needed to assess the outcome and effectiveness of CoQ(10) before exposing patients to unnecessary health risks at significant costs.

Two-dimensional electron states bound to an off-plane donor in a magnetic field

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

On anomalous diffusion and the fractional generalized Langevin equation for a harmonic oscillator

The fractional generalized Langevin equation (FGLE) is proposed to discuss the anomalous diffusive behavior of a harmonic oscillator driven by a two-parameter Mittag-Leffler noise. The solution of this FGLE is discussed by means of the Laplace transform methodology and the kernels are presented in terms of the three-parameter Mittag-Leffler functions. Recent results associated with a generalized Langevin equation are recovered.

Relating pseudospin and spin symmetries through charge conjugation and chiral transformations: the case of the relativistic harmonic oscillator

We solve the generalized relativistic harmonic oscillator in 1+1 dimensions, i.e., including a linear pseudoscalar potential and quadratic scalar and vector potentials which have equal or opposite signs. We consider positive and negative quadratic potentials and discuss in detail their bound-state solutions for fermions and antifermions. The main features of these bound states are the same as the ones of the generalized three-dimensional relativistic harmonic oscillator bound states. The solutions found for zero pseudoscalar potential are related to the spin and pseudospin symmetry of the Dirac equation in 3+1 dimensions. We show how the charge conjugation and gamma(5) chiral transformations relate the several spectra obtained and find that for massless particles the spin and pseudospin symmetry-related problems have the same spectrum but different spinor solutions. Finally, we establish a relation of the solutions found with single-particle states of nuclei described by relativistic mean-field theories with scalar, vector, and isoscalar tensor interactions and discuss the conditions in which one may have both nucleon and antinucleon bound states.

Collider aspects of flavor physics at high Q

This chapter of the "Flavor in the era of LHC" workshop report discusses flavor-related issues in the production and decays of heavy states at the LHC at high momentum transfer Q, both from the experimental and the theoretical perspective. We review top quark physics, and discuss the flavor aspects of several extensions of the standard model, such as supersymmetry, little Higgs models or models with extra dimensions. This includes discovery aspects, as well as the measurement of several properties of these heavy states. We also present publicly available computational tools related to this topic.

Spinless bosons embedded in a vector Duffin-Kemmer-Petiau oscillator

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Estudo comparativo de abordagens de inversão do fator de qualidade Q

Seismic wave dispersion and attenuation studies have become an important tool for lithology and fluid discrimination in hydrocarbon reservoirs. The processes associated to attenuation are complex and are encapsulated in a single quantitative description called quality factor (Q). The present dissertation has the objective of comparing different approaches of Q determination and is divided in two parts. Firstly, we made performance and robustness tests of three different approaches for Q determination in the frequency domain. They are: peak shift, centroid shift and spectral ratio. All these tests were performed in a three-layered model. In the suite of tests performed here, we varied the thickness, Q and inclination of the layers for propagation pulses with central frequency of 30, 40 and 60 Hz. We found that the centroid shift method is produces robust results for the entire suíte of tests. Secondly, we inverted for Q values using the peak and centroid shift methods using an sequential grid search algorithm. In this case, centroid shift method also produced more robust results than the peak shift method, despite being of slower convergence

Expressão gênica dos REα, REβ e PR em tumores mamários de cadelas por meio do q-PCR

A expressão de receptores de estrógeno (ER) e progesterona (PR) por meio da técnica de q-PCR foi avaliada em 26 cadelas portadoras de neoplasias mamárias e cinco cadelas sem afecções mamárias (grupo controle). Os resultados mostraram que os três grupos de animais estudados - com tumor maligno ou benigno e controle - expressaram receptores de estrógeno alfa, beta e progesterona. A quantificação relativa mostrou tendência para uma expressão maior de receptores no grupo controle e menor no grupo de animais com neoplasias malignas. Além disso, observou-se expressão maior de ERα em relação ao ERβ, e as neoplasias malignas de origem mista apresentaram maiores concentrações dos receptores PR, ERα e ERβ que as neoplasias de origem epitelial.

Some experiments with high order compact methods using a computer algebra software - Part I

We consider a procedure for obtaining a compact fourth order method to the steady 2D Navier-Stokes equations in the streamfunction formulation using the computer algebra system Maple. The resulting code is short and from it we obtain the Fortran program for the method. To test the procedure we have solved many cavity-type problems which include one with an analytical solution and the results are compared with results obtained by second order central differences to moderate Reynolds numbers. (c) 2005 Elsevier B.V. All rights reserved.

Some experiments with high order compact methods using a computer algebra software - Part II (non-uniform grid)

We generalize a procedure proposed by Mancera and Hunt [P.F.A. Mancera, R. Hunt, Some experiments with high order compact methods using a computer algebra software-Part 1, Appl. Math. Comput., in press, doi: 10.1016/j.amc.2005.05.015] for obtaining a compact fourth-order method to the steady 2D Navier-Stokes equations in the streamfunction formulation-vorticity using the computer algebra system Maple, which includes conformal mappings and non-uniform grids. To analyse the procedure we have solved a constricted stepped channel problem, where a fine grid is placed near the re-entrant corner by transformation of the independent variables. (c) 2006 Elsevier B.V. All rights reserved.

The Saccharomyces cerevisiae COQ10 gene encodes a START domain protein required for function of coenzyme Q in respiration

Deletion of the Saccharomyces cerevisiae gene YOL008W, here referred to as COQ10, elicits a respiratory defect as a result of the inability of the mutant to oxidize NADH and succinate. Both activities are restored by exogenous coenzyme Q(2). Respiration is also partially rescued by COQ2, COQ7, or COQ8/ABC1, when these genes are present in high copy. Unlike other coq mutants, all of which lack Q(6), the coq10 mutant has near normal amounts of Q(6) in mitochondria. Coq10p is widely distributed in bacteria and eukaryotes and is homologous to proteins of the aromatic-rich protein family Pfam03654 and to members of the START domain superfamily that have a hydrophobic tunnel implicated in binding lipophilic molecules such as cholesterol and polyketides. Analysis of coenzyme Q in polyhistidine-tagged Coq10p purified from mitochondria indicates the presence 0.032-0.034 mol of Q(6)/mol of protein. We propose that Coq10p is a Q(6)-binding protein and that in the coq10 mutant Q(6) it is not able to act as an electron carrier, possibly because of improper localization.

The Q-D algorithm for transforming series expansions into a corresponding continued fraction: an extension to cope with zero coefficients

An algorithm for deriving a continued fraction that corresponds to two series expansions simultaneously, when there are zero coefficients in one or both series, is given. It is based on using the Q-D algorithm to derive the corresponding fraction for two related series, and then transforming it into the required continued fraction. Two examples are given. (C) 2003 Elsevier B.V. All rights reserved.