Optimization of neural classifiers based on bayesian decision boundaries and idle neurons pruning

In this article we describe a feature extraction algorithm for pattern classification based on Bayesian Decision Boundaries and Pruning techniques. The proposed method is capable of optimizing MLP neural classifiers by retaining those neurons in the hidden layer that realy contribute to correct classification. Also in this article we proposed a method which defines a plausible number of neurons in the hidden layer based on the stem-and-leaf graphics of training samples. Experimental investigation reveals the efficiency of the proposed method. © 2002 IEEE.

Precision of yule-walker methods for the arma spectral model

In this work a new method is proposed of separated estimation for the ARMA spectral model based on the modified Yule-Walker equations and on the least squares method. The proposal of the new method consists of performing an AR filtering in the random process generated obtaining a new random estimate, which will reestimate the ARMA model parameters, given a better spectrum estimate. Some numerical examples will be presented in order to ilustrate the performance of the method proposed, which is evaluated by the relative error and the average variation coefficient.

Apparent digestibility of different ingredients in diets for bullfrog Rana catesbeiana tadpoles

The objective of this study was to determine apparent digestibility of protein and energy of ingredients commonly present in rations fed to bullfrog tadpoles Rana catesbeiana such as corn, soybean meal, wheat meal, rice bran, poultry by-product meal, blood meal, fish meal, corn starch and soybean oil. A completely randomized design was used to test ten experimental diets with four replicates. About 30% of the reference diet was replaced by the test ingredients, except for soybean oil that was used at 10% ratio and 0.5% of chromium oxide as inert marker. The values of apparent digestibility coefficient of crude protein (ADCCP) and gross energy (ADCGE) of the protein ingredients were, respectively: fish meal, 81 and 95%; poultry byproduct meal, 82 and 87%; blood meal, 64 and 53%; and soybean meal, 93 and 90%. While for the energetic ingredients, ADCCP and ADCGE, were, respectively: wheat meal, 100 and 72%; rice bran, 57 and 24%; and corn, 84 and 92%. Soybean oil and cornstarch ADCGE values were 70 and 96%, respectively; cornstarch digestibility is highly dependent on feed processing technique. The results showed that the ing redients of plant origin, except for rice bran, had good digestibility, thus indicating that tadpoles can effectively use plant protein. In this trial, the least digestible ingredients were blood meal and rice bran, thus indicating that the use of such ingredients in bullfrog tadpole diets should be limited. © Copyright by the World Aquaculture Society 2005.

Avaliação estatística de métodos de poda aplicados em neurônios intermediários da rede neural MLP

There are several papers on pruning methods in the artificial neural networks area. However, with rare exceptions, none of them presents an appropriate statistical evaluation of such methods. In this article, we proved statistically the ability of some methods to reduce the number of neurons of the hidden layer of a multilayer perceptron neural network (MLP), and to maintain the same landing of classification error of the initial net. They are evaluated seven pruning methods. The experimental investigation was accomplished on five groups of generated data and in two groups of real data. Three variables were accompanied in the study: apparent classification error rate in the test group (REA); number of hidden neurons, obtained after the application of the pruning method; and number of training/retraining epochs, to evaluate the computational effort. The non-parametric Friedman's test was used to do the statistical analysis.

Prospective analysis about the world wide strategy of commercial vehicles development

The main objective of this paper is to make a prospective analysis about the tendencies of the world wide strategies of commercial vehicles development, in relation of the current local and international scenery. It has been done a research by using the Delphi method, which consists into regulate, systematically, the presentation of questions about a specified problem to a group of specialists. In order to guarantee more representative data, it has been adopted three distinct groups: academics, OEMs managers and Tier 1 suppliers managers. The results showed that the three mentioned groups have similar opinion about the necessity of having a product development in the subsidiaries companies; also it has been pointed the necessity of having more taylor made vehicles, due to the intrinsic characteristics of this industry segment. Copyright © 2008 SAE International.

Estudo da perda da tensao do enxerto de tendão calcâneo bovino

Lesion of the anterior cruciate ligament (ACL) in the knee is a problem that affects animals as well as humans and may evolve with joint instability that is often symptomatic and incapacitating. The main option for ligament substitution is the autologous tendon graft. The graft undergoes tensioning during ACL reconstructive surgery to reestablish the normal laxity of the ACL- deficient knee. Although graft tensioning plays a fundamental role in postoperative clinical evolution, ideal tensioning levels have not been established in the literature. Therefore, graft elongation that may occur over time is still one of the main reasons for ligament reconstruction failure. In this study, ten bovine calcaneus tendons underwent two successive assays of physiological tensioning for a maximum deformation of 2.5% of the initial tendon length, maintained for 600s with force (N) values recorded at zero time (initial), 300s and 600s. At the end of the first assay, the tendon returned to its initial length and was maintained at rest for 300s and then the next tensioning assay was initiated, repeating the previous procedure. Statistical analysis revealed that tendon elongation is more pronounced during the initial 300s, reflecting a more accentuated decline in the tension values of the graft. Therefore, 300s after tensioning and fixing the graft, the surgeon can assess with more precision if the tensioning level was adequate.

Comparison of GEANT4 simulations with experimental data for thick al absorbers

Proton beams in medical applications deal with relatively thick targets like the human head or trunk. Therefore, relatively small differences in the total proton stopping power given, for example, by the different models provided by GEANT4 can lead to significant disagreements in the final proton energy spectra when integrated along lengthy proton trajectories. This work presents proton energy spectra obtained by GEANT4.8.2 simulations using ICRU49, Ziegler1985 and Ziegler2000 models for 19.68MeV protons passing through a number of Al absorbers with various thicknesses. The spectra were compared with the experimental data, with TRIM/SRIM2008 and MCNPX2.4.0 simulations, and with the Payne analytical solution for the transport equation in the Fokker-Plank approximation. It is shown that the MCNPX simulations reasonably reproduce well all experimental spectra. For the relatively thin targets all the methods give practically identical results but this is not the same for the thick absorbers. It should be noted that all the spectra were measured at the proton energies significantly above 2MeV, i.e., in the so-called Bethe-Bloch region. Therefore the observed disagreements in GEANT4 results, simulated with different models, are somewhat unexpected. Further studies are necessary for better understanding and definitive conclusions. © 2009 American Institute of Physics.

Reduced calibration curve for proton computed tomography

The pCT deals with relatively thick targets like the human head or trunk. Thus, the fidelity of pCT as a tool for proton therapy planning depends on the accuracy of physical formulas used for proton interaction with thick absorbers. Although the actual overall accuracy of the proton stopping power in the Bethe-Bloch domain is about 1%, the analytical calculations and the Monte Carlo simulations with codes like TRIM/SRIM, MCNPX and GEANT4 do not agreed with each other. A tentative to validate the codes against experimental data for thick absorbers bring some difficulties: only a few data is available and the existing data sets have been acquired at different initial proton energies, and for different absorber materials. In this work we compare the results of our Monte Carlo simulations with existing experimental data in terms of reduced calibration curve, i.e. the range - energy dependence normalized on the range scale by the full projected CSDA range for given initial proton energy in a given material, taken from the NIST PSTAR database, and on the final proton energy scale - by the given initial energy of protons. This approach is almost energy and material independent. The results of our analysis are important for pCT development because the contradictions observed at arbitrary low initial proton energies could be easily scaled now to typical pCT energies. © 2010 American Institute of Physics.

Prompt and non-prompt J/ψ production in pp collisions at √s =7 TeV

The production of J/ψ mesons is studied in pp collisions at √s = 7 TeV with the CMS experiment at the LHC. The measurement is based on a dimuon sample corresponding to an integrated luminosity of 314 nb-1. The J/ψ differential cross section is determined, as a function of the J/ψ transverse momentum, in three rapidity ranges. A fit to the decay length distribution is used to separate the prompt from the non-prompt (b hadron to J/ψ) component. Integrated over J/ψ transverse momentum from 6.5 to 30 GeV/c and over rapidity in the range {pipe}y{pipe} < 2.4, the measured cross sections, times the dimuon decay branching fraction, are 70.9 ± 2.1(stat.) ± 3.0(syst.) ± 7.8(luminosity) nb for prompt J/ψ mesons assuming unpolarized production and 26.0 ± 1.4(stat.) ± 1.6(syst.) ± 2.9(luminosity) nb for J/ψ mesons from b-hadron decays. © CERN for the benefit of the CMS collaboration 2011.

GEANT4 tuning for pCT development

Proton beams in medical applications deal with relatively thick targets like the human head or trunk. Thus, the fidelity of proton computed tomography (pCT) simulations as a tool for proton therapy planning depends in the general case on the accuracy of results obtained for the proton interaction with thick absorbers. GEANT4 simulations of proton energy spectra after passing thick absorbers do not agree well with existing experimental data, as showed previously. Moreover, the spectra simulated for the Bethe-Bloch domain showed an unexpected sensitivity to the choice of low-energy electromagnetic models during the code execution. These observations were done with the GEANT4 version 8.2 during our simulations for pCT. This work describes in more details the simulations of the proton passage through aluminum absorbers with varied thickness. The simulations were done by modifying only the geometry in the Hadrontherapy Example, and for all available choices of the Electromagnetic Physics Models. As the most probable reasons for these effects is some specific feature in the code, or some specific implicit parameters in the GEANT4 manual, we continued our study with version 9.2 of the code. Some improvements in comparison with our previous results were obtained. The simulations were performed considering further applications for pCT development. © 2011 American Institute of Physics.

Long-range and short-range dihadron angular correlations in central PbPb collisions at psNN = 2:76 TeV

First measurements of dihadron correlations for charged particles are presented for central PbPb collisions at a nucleon-nucleon center-of-mass energy of 2.76TeV over a broad range in relative pseudorapidity (Δν) and the full range of relative azimuthal angle (Δø). The data were collected with the CMS detector, at the LHC. A broadening of the away-side (Δø y≈ π) azimuthal correlation is observed at all Δν, as compared to the measurements in pp collisions. Furthermore, long-range dihadron correlations in Δν are observed for particles with similar ø values. This phenomenon, also known as the \ridge, persists up to at least jΔνj = 4. For particles with transverse momenta (pT) of 2-4 GeV/c, the ridge is found to be most prominent when these particles are correlated with particles of pT = 2-6 GeV/c, and to be much reduced when paired with particles of pT = 10-12 GeV/c. Copyright CERN.

Measurements of inclusive W and Z cross sections in pp collisions at √s = 7 TeV The CMS collaboration

Measurements of inclusive W and Z boson production cross sections in pp collisions at √s = 7 TeV are presented, based on 2.9 pb-1 of data recorded by the CMS detector at the LHC. The measurements, performed in the electron and muon decay channels, are combined to give σ(pp → WX) × B(W → l?) = 9.95 ± 0.07 (stat.) ± 0.28 (syst.) ± 1.09 (lumi.) nb and σ(pp → ZX) × B(Z → l +l-) = 0.931 ± 0.026 (stat.) ± 0.023 (syst.) ± 0.102 (lumi.) nb, where l stands for either e or μ. Theoretical predictions, calculated at the next-to-next-to-leading order in QCD using recent parton distribution functions, are in agreement with the measured cross sections. Ratios of cross sections, which incur an experimental systematic uncertainty of less than 4%, are also reported.

Measurement of the Drell-Yan cross section in pp collisions at √ s = 7 TeV

The Drell-Yan differential cross section is measured in pp collisions at √ s = 7TeV, from a data sample collected with the CMS detector at the LHC, corresponding to an integrated luminosity of 36 pb-1. The cross section measurement, normalized to the measured cross section in the Z region, is reported for both the dimuon and dielectron channels in the dilepton invariant mass range 15{600 GeV. The normalized cross section values are quoted both in the full phase space and within the detector acceptance. The effect of final state radiation is also identified. The results are found to agree with theoretical predictions. Copyright CERN.

Charged particle transverse momentum spectra in pp collisions at √s = 0:9 and 7 TeV

The charged particle transverse momentum (pT) spectra are presented for pp collisions at √s = 0:9 and 7TeV. The data samples were collected with the CMS detector at the LHC and correspond to integrated luminosities of 231 μb-1 and 2.96 pb-1, respectively. Calorimeter-based high-transverse-energy triggers are employed to enhance the statistical reach of the high-pT measurements. The results are compared with leading and next-toleading order QCD and with an empirical scaling of measurements at different collision energies using the scaling variable xT - 2pT=ps over the pT range up to 136 GeV/c. Using a combination of xT scaling and direct interpolation at fixed pT, a reference transverse momentum spectrum at √s = 2:76TeV is constructed, which can be used for studying high-pT particle suppression in the dense QCD medium produced in heavy-ion collisions at that centre-of-mass energy. Copyright CERN.

Search for supersymmetry in pp collisions at √s = 7 TeV in events with a single lepton, jets, and missing transverse momentum

Results are reported from a search for physics beyond the standard model in proton-proton collisions at a center-of-mass energy of 7TeV, focusing on the signature with a single, isolated, high-transverse-momentum lepton (electron or muon), energetic jets, and large missing transverse momentum. The data sample comprises an integrated luminosity of 36 pb-1, recorded by the CMS experiment at the LHC. The search is motivated by models of new physics, including supersymmetry. The observed event yields are consistent with standard model backgrounds predicted using control samples obtained from the data. The characteristics of the event sample are consistent with those expected for the production of tt̄ and W+jets events. The results are interpreted in terms of limits on the parameter space for the constrained minimal supersymmetric extension of the standard model.