Activation function study for wavelet network

The main purpose of this paper is to investigate theoretically and experimentally the use of family of Polynomial Powers of the Sigmoid (PPS) Function Networks applied in speech signal representation and function approximation. This paper carries out practical investigations in terms of approximation fitness (LSE), time consuming (CPU Time), computational complexity (FLOP) and representation power (Number of Activation Function) for different PPS activation functions. We expected that different activation functions can provide performance variations and further investigations will guide us towards a class of mappings associating the best activation function to solve a class of problems under certain criteria.

Análise de pathway e network em tecido adiposo in vivo e o papel da vitamina D na adipogênese in vitro

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Mathematical tests about the existence and applications of PPS-wavelets in function approximation problems

Neural networks and wavelet transform have been recently seen as attractive tools for developing eficient solutions for many real world problems in function approximation. Function approximation is a very important task in environments where computation has to be based on extracting information from data samples in real world processes. So, mathematical model is a very important tool to guarantee the development of the neural network area. In this article we will introduce one series of mathematical demonstrations that guarantee the wavelets properties for the PPS functions. As application, we will show the use of PPS-wavelets in pattern recognition problems of handwritten digit through function approximation techniques.

An efficient Hopfield network to solve economic dispatch problems with transmission system representation

Economic dispatch (ED) problems have recently been solved by artificial neural network approaches. Systems based on artificial neural networks have high computational rates due to the use of a massive number of simple processing elements and the high degree of connectivity between these elements. The ability of neural networks to realize some complex non-linear function makes them attractive for system optimization. All ED models solved by neural approaches described in the literature fail to represent the transmission system. Therefore, such procedures may calculate dispatch policies, which do not take into account important active power constraints. Another drawback pointed out in the literature is that some of the neural approaches fail to converge efficiently toward feasible equilibrium points. A modified Hopfield approach designed to solve ED problems with transmission system representation is presented in this paper. The transmission system is represented through linear load flow equations and constraints on active power flows. The internal parameters of such modified Hopfield networks are computed using the valid-subspace technique. These parameters guarantee the network convergence to feasible equilibrium points, which represent the solution for the ED problem. Simulation results and a sensitivity analysis involving IEEE 14-bus test system are presented to illustrate efficiency of the proposed approach. (C) 2004 Elsevier Ltd. All rights reserved.

Security constrained optimal active power flow via network model and interior point method

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

Electrical load forecasting formulation by a fast neural network

The objective of this work is to develop a methodology for electric load forecasting based on a neural network. Here, backpropagation algorithm is used with an adaptive process that based on fuzzy logic and using a decaying exponential function to avoid instability in the convergence process. This methodology results in fast training, when compared to the conventional formulation of backpropagation algorithm. The results are presented using data from a Brazilian Electric Company, and shows a very good performance for the proposal objective.

An Efficient Codification to Solve Distribution Network Reconfiguration for Loss Reduction Problem

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Structural Basis of Importin-alpha-Mediated Nuclear Transport for Ku70 and Ku80

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

Structural basis for metal ion coordination and the catalytic mechanism of sphingomyelinases D

Sphingomyelinases D (SMases D) from Loxosceles spider venom are the principal toxins responsible for the manifestation of dermonecrosis, intravascular hemolysis, and acute renal failure, which can result in death. These enzymes catalyze the hydrolysis of sphingomyelin, resulting in the formation of ceramide 1-phosphate and choline or the hydrolysis of lysophosphatidyl choline, generating the lipid mediator lysophosphatidic acid. This report represents the first crystal structure of a member of the sphingomyelinase D family from Loxosceles laeta (SMase I), which has been determined at 1.75-angstrom resolution using the quick cryo-soaking technique and phases obtained from a single iodine derivative and data collected from a conventional rotating anode x-ray source. SMase I folds as an (alpha/beta)(8) barrel, the interfacial and catalytic sites encompass hydrophobic loops and a negatively charged surface. Substrate binding and/or the transition state are stabilized by a Mg2+ ion, which is coordinated by Glu(32), Asp(34), Asp(91), and solvent molecules. In the proposed acid base catalytic mechanism, His(12) and His(47) play key roles and are supported by a network of hydrogen bonds between Asp(34), Asp(52), Trp(230), Asp(233), and Asn(252).

SMase II, a new sphingomyelinase D from Loxosceles laeta venom gland: Molecular cloning, expression, function and structural analysis

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

Structural basis for branching-enzyme activity of glycoside hydrolase family 57: Structure and stability studies of a novel branching enzyme from the hyperthermophilic archaeon Thermococcus Kodakaraensis KOD1

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

On the Baker-Akhiezer function in the AKNS scheme

Expressions for the Baker-Akhiezer function and their logarithmic space and time derivatives are derived in terms of the matrix elements of U - V matrices and 'squared basis functions'. These expressions generalize the well known formulas for the KdV equation case and establish links between different forms of the Whitham averaging procedure.

Myocyte necrosis is the basis for fibrosis inrenovascular hypertensive rats

The pathogenesis of fibrosis and the functional features of pressure overload myocardial hypertrophy are still controversial. The objectives of the present study were to evaluate the function and morphology of the hypertrophied myocardium in renovascular hypertensive (RHT) rats. Male Wistar rats were sacrificed at week 4 (RHT4) and 8 (RHT8) after unilateral renal ischemia (Goldblatt II hypertension model). Normotensive rats were used as controls. Myocardial function was analyzed in isolated papillary muscle preparations, morphological features were defined by light microscopy, and myocardial hydroxyproline concentration (HOP) was determined by spectrophotometry. Renal artery clipping resulted in elevated systolic arterial pressure (RHT4: 178 ± 19 mmHg and RHT8: 194 ± 24 mmHg, P<0.05 vs control: 123 ± 7 mmHg). Myocardial hypertrophy was observed in both renovascular hypertensive groups. The myocardial HOP concentration was increased in the RHT8 group (control: 2.93 ± 0.38 µg/mg; RHT4: 3.02 ± 0.40 µg/mg; RHT8: 3.44 ± 0.45 µg/mg of dry tissue, P<0.05 vs control and RHT4 groups). The morphological study demonstrated myocyte necrosis, vascular damage and cellular inflammatory response throughout the experimental period. The increased cellularity was more intense in the adventitia of the arterioles. As a consequence of myocyte necrosis, there was an early, local, conjunctive stroma collapse with disarray and thickening of the argyrophilic interstitial fibers, followed by scarring. The functional data showed an increased passive myocardial stiffness in the RHT4 group. We conclude that renovascular hypertension induces myocyte and arteriole necrosis. Reparative fibrosis occurred as a consequence of the inflammatory response to necrosis. The mechanical behavior of the isolated papillary muscle was normal, except for an early increased myocardial passive stiffness

GCHF basis sets and their application in the electronic structure study of PrMnO3

The approach called generator coordinate Hartree-Fock (GCHF) method is used in the selection of Gaussian basis set [25s18p for O ((3)p), 31s21p14d for Mn (S-6), and 33s22p16d9f for Pr ((4)J)] for atoms. The role of the weight functions in the assessment of the numerical integration range of the GCHF equations is shown. These basis sets are contracted to (25s18p/9s5p), (31s21p14d/9s6p4d), and (33s22pl6d9f118sl2p5d3f) by segmented contraction scheme of Dunning and they are utilized in calculations of Restricted-Open-HF (ROHF) Total and Orbital energies of the (MnO+1)-Mn-3 and (PrO+1)-Pr-1 fragments, to evaluate their quality in molecular studies. The addition of one d polarization function in the contracted (9s5p) basis set for O(P-3) atom and their application with the contracted (9s6p4d), (18s21p5d3f) basis sets for Mn (S-6) and Pr-Pr ((4)j) atoms lead to the electronic structure study of PrMnO3. The dipole moment, the total energy, and total atomic charges properties were calculated and were carried out at ROHF level with the [PrMnO3](2) fragment. The calculated values show that PrMnO3 does not present piezoelectric properties. (C) 2004 Elsevier B.V. All rights reserved.

Short term hydroelectric scheduling combining network flow and interior point approaches

In this paper, short term hydroelectric scheduling is formulated as a network flow optimization model and solved by interior point methods. The primal-dual and predictor-corrector versions of such interior point methods are developed and the resulting matrix structure is explored. This structure leads to very fast iterations since it avoids computation and factorization of impedance matrices. For each time interval, the linear algebra reduces to the solution of two linear systems, either to the number of buses or to the number of independent loops. Either matrix is invariant and can be factored off-line. As a consequence of such matrix manipulations, a linear system which changes at each iteration has to be solved, although its size is reduced to the number of generating units and is not a function of time intervals. These methods were applied to IEEE and Brazilian power systems, and numerical results were obtained using a MATLAB implementation. Both interior point methods proved to be robust and achieved fast convergence for all instances tested. (C) 2004 Elsevier Ltd. All rights reserved.