Theoretical foundations of neural networks

Neural networks have often been motivated by superficial analogy with biological nervous systems. Recently, however, it has become widely recognised that the effective application of neural networks requires instead a deeper understanding of the theoretical foundations of these models. Insight into neural networks comes from a number of fields including statistical pattern recognition, computational learning theory, statistics, information geometry and statistical mechanics. As an illustration of the importance of understanding the theoretical basis for neural network models, we consider their application to the solution of multi-valued inverse problems. We show how a naive application of the standard least-squares approach can lead to very poor results, and how an appreciation of the underlying statistical goals of the modelling process allows the development of a more general and more powerful formalism which can tackle the problem of multi-modality.

Rugosidade e pigmentação superficial de materiais ionoméricos

The aim of this study was to compare two methods of surface roughness analysis, perfilometry and spectrophotometry, applied to the surface of ionomeric materials (Chelon Fil, Vitremer and Dyract), submitted to different surface finishing treatments. For the perfilometric analysis, sixty specimens of each material were made and randomly separated into three experimental groups. The average surface roughness (Ra, mm) was measured on each specimen by a surface perfilometer (Mitutoyo Surftest 211). The spectrophotometric analysis consisted in quantifying the dye impregnated in the samples. The dyes used were 0.5% fuchsin and 0.5% erythrosin. Data were submitted to variance analysis (ANOVA) and t-Student test at a 0.05 significance level. There was no linear correlation between average roughness and superficial deposition of dye. Perfilometric analysis revealed that 12- and 30-bladed carbide burs caused the roughest surface of Chelon Fil, followed by Sof-Lex discs and mylar band. There were no significant differences between the specimens submitted to finishing and polishing with Sof-Lex discs and the control group (mylar band) for Vitremer, nevertheless, the highest Ra values were obtained when 12- and 30-bladed burs were used. For Dyract, there was no significant difference between the three treatments. The mean values of superficial deposition of dye for Chelon Fil, Vitremer and Dyract were: 1.7261, 1.4759, 1.3318, respectively. There were no significant differences between the restorative materials when different finishing and polishing systems were used.

Evaluation of superficial microhardness in dental enamel with different eruptive ages

This study evaluated the superficial microhardness of enamel in teeth at different posteruptive ages (before eruption in the oral cavity, 2-3 years after eruption, 4-10 years after eruption and more than 10 years after eruption). The study sample was composed of 134 specimens of human enamel. One fragment of each tooth was obtained from the flattest central portion of the crown to produce specimens with 3 x 3 mm. The enamel blocks were minimally flattened out and polished in order to obtain a flat surface parallel to the base, which is fundamental for microhardness testing. Microhardness was measured with a microhardness tester and a Knoop diamond indenter, under a static load of 25 g applied for 5 seconds. Comparison between the superficial microhardness obtained for the different groups was performed by analysis of Student's t test. The results demonstrated that superficial microhardness values have a tendency to increase over the years, with statistically significant difference only between unerupted enamel and that with more than 10 years after eruption. According to the present conditions and methodology, it was concluded that there were differences between the superficial micro-hardness of specimens at different eruptive ages, revealing an increasing mineralization. However, this difference was significant only between unerupted specimens and those with more than 10 years after eruption.

High-energy events and short-term changes in superficial beach sediments

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Disponibilidade de metais pesados em Latossolo com aplicação superficial de escória, lama cal, lodos de esgoto e calcário

O objetivo deste trabalho foi avaliar alguns atributos químicos do solo e a disponibilidade de cádmio (Cd), cromo (Cr), níquel (Ni), mercúrio (Hg), chumbo (Pb) e arsênio (As), por meio da extração pelo DTPA,em conseqüência da aplicação superficial de escória de aciaria, lama cal e lodos de esgoto centrifugados e de biodigestores, nas doses 0 (testemunha), 2, 4 e 8 Mg ha-1 e um tratamento adicional composto pela calagem superficial na dose 2 Mg ha-1. O experimento foi conduzido em delineamento de blocos ao acaso, em condições de campo, em área sob sistema plantio direto, durante 2003 e 2004. A aplicação superficial de escória de aciaria,lama cal, lodo de esgoto centrifugado e de biodigestor, até a dose 8 Mg ha-1, assim como o calcário na dose 2 Mg ha-1, não trazem problemas de disponibilidade ao ambiente, com relação aos metais pesados Cd, Cr, Hg, Pb, Ni e As, quando aplicados sobre a superfície em Latossolo Vermelho distrófico, no sistema plantio direto. A fitodisponibilidade de metais pesados às culturas da soja e aveia-preta foi nula, quando foram aplicadas doses de até 8 Mg ha-1 de lodo de esgoto, escória e lama cal sobre a superfície do solo, no sistema plantio direto

Determination of Threshold Dose of Photodynamic Therapy to Measure Superficial Necrosis

Background Data: Photodynamic therapy (PDT) involves the photoinduction of cytotoxicity using a photosensitizer agent, a light source of the proper wavelength, and the presence of molecular oxygen. A model for tissue response to PDT based on the photodynamic threshold dose (Dth) has been widely used. In this model cells exposed to doses below Dth survive while at doses above the Dth necrosis takes place. Objective: This study evaluated the light Dth values by using two different methods of determination. One model concerns the depth of necrosis and the other the width of superficial necrosis. Materials and Methods: Using normal rat liver we investigated the depth and width of necrosis induced by PDT when a laser with a gaussian intensity profile is used. Different light doses, photosensitizers (Photogem, Photofrin, Photosan, Foscan, Photodithazine, and Radachlorin), and concentrations were employed. Each experiment was performed on five animals and the average and standard deviations were calculated. Results: A simple depth and width of necrosis model analysis allows us to determine the threshold dose by measuring both depth and surface data. Comparison shows that both measurements provide the same value within the degree of experimental error. Conclusion: This work demonstrates that by knowing the extent of the superficial necrotic area of a target tissue irradiated by a gaussian light beam, it is possible to estimate the threshold dose. This technique may find application where the determination of Dth must be done without cutting the tissue.

Ablation Rate and Morphology of Superficial and Deep Dentin Irradiated with Different Er:YAG Laser Energy Levels

Objective: In this study we evaluated the ablation rate of superficial and deep dentin irradiated with different Er:YAG laser energy levels, and observed the micromorphological aspects of the lased substrates with a scanning electron microscope (SEM). Background Data: Little is known about the effect of Er: YAG laser irradiation on different dentin depths. Materials and Methods: Sixty molar crowns were bisected, providing 120 specimens, which were randomly assigned into two groups ( superficial or deep dentin), and later into five subgroups (160, 200, 260, 300, or 360 mJ). Initial masses of the specimens were obtained. After laser irradiation, the final masses were obtained and mass losses were calculated followed by the preparation of specimens for SEM examination. Mass-loss values were subjected to two-way ANOVA and Fisher's least significant difference multiple-comparison tests (p < 0.05). Results: There was no difference between superficial and deep dentin. A significant and gradual increase in the mass-loss values was reached when energies were raised, regardless of the dentin depth. The energy level of 360 mJ showed the highest values and was statistically significantly different from the other energy levels. The SEM images showed that deep dentin was more selectively ablated, especially intertubular dentin, promoting tubule protrusion. At 360 mJ the micromorphological features were similar for both dentin depths. Conclusion: The ablation rate did not depend on the depth of the dentin, and an energy level lower than 360 mJ is recommended to ablate both superficial and deep dentin effectively without causing tissue damage.

Bending of stochastic Kirchhoff plates on Winkler foundations via the Galerkin method and the Askey-Wiener scheme

In this paper, the method of Galerkin and the Askey-Wiener scheme are used to obtain approximate solutions to the stochastic displacement response of Kirchhoff plates with uncertain parameters. Theoretical and numerical results are presented. The Lax-Milgram lemma is used to express the conditions for existence and uniqueness of the solution. Uncertainties in plate and foundation stiffness are modeled by respecting these conditions, hence using Legendre polynomials indexed in uniform random variables. The space of approximate solutions is built using results of density between the space of continuous functions and Sobolev spaces. Approximate Galerkin solutions are compared with results of Monte Carlo simulation, in terms of first and second order moments and in terms of histograms of the displacement response. Numerical results for two example problems show very fast convergence to the exact solution, at excellent accuracies. The Askey-Wiener Galerkin scheme developed herein is able to reproduce the histogram of the displacement response. The scheme is shown to be a theoretically sound and efficient method for the solution of stochastic problems in engineering. (C) 2009 Elsevier Ltd. All rights reserved.

Galerkin Solution of Stochastic Beam Bending on Winkler Foundations

In this paper, the Askey-Wiener scheme and the Galerkin method are used to obtain approximate solutions to stochastic beam bending on Winkler foundation. The study addresses Euler-Bernoulli beams with uncertainty in the bending stiffness modulus and in the stiffness of the foundation. Uncertainties are represented by parameterized stochastic processes. The random behavior of beam response is modeled using the Askey-Wiener scheme. One contribution of the paper is a sketch of proof of existence and uniqueness of the solution to problems involving fourth order operators applied to random fields. From the approximate Galerkin solution, expected value and variance of beam displacement responses are derived, and compared with corresponding estimates obtained via Monte Carlo simulation. Results show very fast convergence and excellent accuracies in comparison to Monte Carlo simulation. The Askey-Wiener Galerkin scheme presented herein is shown to be a theoretically solid and numerically efficient method for the solution of stochastic problems in engineering.

Analysing the base of precast column in socket foundations with smooth interfaces

This paper analyzes the behavior of the base of a precast column in the socket foundation with smooth interfaces. This research is motivated by the lack of information and guidelines on the behavior of column bases in the embedded region. An experimental program with two full-scale specimens was carried-out. These two specimens had smooth interfaces at the internal faces of the socket, different embedded lengths and were subjected to loads with large eccentricities. The experimental results showed that the failure of the specimens occurred by the yielding of the longitudinal reinforcement out of the embedded region, while the transverse reinforcement was not very stressed. Some recommendations on the anchorage of the longitudinal reinforcement and a strut-and-tie model for the behavior of column bases in the embedded region are proposed.

The synergism between high superficial roughness and corrosive atmosphere in the fatigue behavior of the SAE 1045 steel

The fatigue behavior of the SAE 1045 steel in four different conditions is reported in this paper. Four treatments were tested: polished surface, corrosive atmosphere (sugar cane juice), low quality roughness (typical equivalent surface of the shaft cane mill) and the combined condition of the two last treatments involved. Fatigue tests were performed in rotating bending conditions and the values of fatigue strength for the different situations from tests were obtained. The diminution in fatigue strength for the synergy condition of rough surface and corrosive atmosphere was of 8.7%, result statistically representative with less than 0.01% of uncertainty.

Public policy in knowledge-based economies: Foundations and frameworks

Knowledge is a product of human social systems and, therefore, the foundations of the knowledge-based economy are social and cultural. Communication is central to knowledge creation and diffusion, and Public Policy in Knowledge-Based Economies highlights specific social and cultural conditions that can enhance the communication, use and creation of knowledge in a society.The purpose of this book is to illustrate how these social and cultural conditions are identified and analysed through new conceptual frameworks. Such frameworks are necessary to penetrate the surface features of knowledge-based economies - science and technology - and disclose what drives such economies.This book will provide policymakers, analysts and academics with the fundamental tools needed for the development of policy in this little understood and emerging area.