An adaptive system based on artificial neural networks for NPK fluid fertilizer application

The application process of fluid fertilizers through variable rates implemented by classical techniques with feedback and conventional equipments can be inefficient or unstable. This paper proposes an open-loop control system based on artificial neural network of the type multilayer perceptron for the identification and control of the fertilizer flow rate. The network training is made by the algorithm of Levenberg-Marquardt with training data obtained from measurements. Preliminary results indicate a fast, stable and low cost control system for precision fanning. Copyright (C) 2000 IFAC.

A Non-Destructive Testing Based on Electromagnetic Measurements and Neural Networks for the Inspection of Concrete Structures

This paper presents a new non-destructive testing (NDT) for reinforced concrete structures, in order to identify the components of their reinforcement. A time varying electromagnetic field is generated close to the structure by electromagnetic devices specially designed for this purpose. The presence of ferromagnetic materials (the steel bars of the reinforcement) immersed in the concrete disturbs the magnetic field at the surface of the structure. These field alterations are detected by sensors coils placed on the concrete surface. Variations in position and cross section (the size) of steel bars immersed in concrete originate slightly different values for the induced voltages at the coils.. The values for the induced voltages were obtained in laboratory tests, and multi-layer perceptron artificial neural networks with Levemberg-Marquardt training algorithm were used to identify the location and size of the bar. Preliminary results can be considered very good.

Training courses for teachers through the distance learning: practice reports used virtual environment TELEDUC

This paper describes the experiences of long-distance courses, it focused on the continuing education of basic education teachers in all Brazilian territory. Such courses were offered by CECEMCA (Center for Continuing Education in Mathematics Education, Science and Environment), linked to the Universidade Estadual Paulista Julio de Mesquita Filho - Campus de Rio Claro during 15/01/2009 to 30/11/2009. The subjects report to the theme of Education, Geography and Environment, it was organized in four courses: "Introduction to Cartography," Environment and climate change - thinking a new paradigm of sustainable green planet "," Remote Sensing in environmental studies Environment "and" Methodological Alternatives for Inclusive Classroom: Experimenting with visual and hearing impairments". So, we show here, the feasibility and importance of distance learning tools for education, specifically teacher training, based on the results obtained in these courses.

The workload and plasma ion concentration in a training match session of high-goal (elite) polo ponies

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

Neural networks training using the constructivism paradigms

The Backpropagation Algorithm (BA) is the standard method for training multilayer Artificial Neural Networks (ANN), although it converges very slowly and can stop in a local minimum. We present a new method for neural network training using the BA inspired on constructivism, an alphabetization method proposed by Emilia Ferreiro based on Piaget philosophy. Simulation results show that the proposed configuration usually obtains a lower final mean square error, when compared with the standard BA and with the BA with momentum factor.

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.

Transient stability analysis of electrical power systems using a neural network based on fuzzy ARTMAP

This work presents a methodology to analyze transient stability for electric energy systems using artificial neural networks based on fuzzy ARTMAP architecture. This architecture seeks exploring similarity with computational concepts on fuzzy set theory and ART (Adaptive Resonance Theory) neural network. The ART architectures show plasticity and stability characteristics, which are essential qualities to provide the training and to execute the analysis. Therefore, it is used a very fast training, when compared to the conventional backpropagation algorithm formulation. Consequently, the analysis becomes more competitive, compared to the principal methods found in the specialized literature. Results considering a system composed of 45 buses, 72 transmission lines and 10 synchronous machines are presented. © 2003 IEEE.

Exercise training in the aerobic/anaerobic metabolic transition prevents glucose intolerance in alloxan-treated rats

Background: Ninety percent of cases of diabetes are of the slowly evolving non-insulin-dependent type, or Type 2 diabetes. Lack of exercise is regarded as one of the main causes of this disorder. In this study we analyzed the effects of physical exercise on glucose homeostasis in adult rats with type 2 diabetes induced by a neonatal injection of alloxan. Methods: Female Wistar rats aged 6 days were injected with either 250 mg/ kg of body weight of alloxan or citrate buffer 0.01 M (controls). After weaning, half of the animals in each group were subjected to physical training adjusted to meet the aerobic-anaerobic metabolic transition by swimming 1 h/day for 5 days a week with weight overloads. The necessary overload used was set and periodically readjusted for each rat through effort tests based on the maximal lactate steady state procedure. When aged 28, 60, 90, and 120 days, the rats underwent glucose tolerance tests (GTT) and their peripheral insulin sensitivity was evaluated using the HOMA index. Results: The area under the serum glucose curve obtained through GTT was always higher in alloxan-treated animals than in controls. A decrease in this area was observed in trained alloxan-treated rats at 90 and 120 days old compared with non-trained animals. At 90 days old the trained controls showed lower HOMA indices than the non-trained controls. Conclusion: Neonatal administration of alloxan induced a persistent glucose intolerance in all injected rats, which was successfully counteracted by physical training in the aerobic/anaerobic metabolic transition. © 2008 Mota et al; licensee BioMed Central Ltd.

Possible shunt compensation strategies based on conservative power theory

Considering the Conservative Power Theory (CPT), this paper proposes some novel compensation strategies for shunt passive or active devices. The CPT current decompositions result in several current terms, which are associated with specific physical phenomena (average power consumption P, energy storage Q, load and source distortion D, unbalances N). These current components were used in this work for the definition of different current compensators, which can be selective in terms of minimizing particular disturbing effects. Compensation strategies for single and three-phase four-wire circuits have also been considered. Simulation results have been demonstrated in order to validate the possibilities and performance of the proposed strategies. © 2010 IEEE.

On the determination of epsilon during discriminative GMM training

Discriminative training of Gaussian Mixture Models (GMMs) for speech or speaker recognition purposes is usually based on the gradient descent method, in which the iteration step-size, ε, uses to be defined experimentally. In this letter, we derive an equation to adaptively determine ε, by showing that the second-order Newton-Raphson iterative method to find roots of equations is equivalent to the gradient descent algorithm. © 2010 IEEE.

Responses of blood lactate concentration in aerobic and anaerobic training protocols at different swimming exercise intensities in rats

A swimming periodized experimental training model in rats in which different training protocols (TP) were classified in aerobic (A) and anaerobic (AN) intensity levels. The purpose of the present study was to verify if the classification of the TP used in the periodized training experimental model presented the blood lactate concentration [La] response adequate to the aerobic and anaerobic intensities levels. Twenty three male Wistar rats were divided into three groups. Two groups of swimming training (continuous, CT, n = 7, and periodized training, PET, n = 7) rats were evaluated during 5 weeks in eight different TP (TP-1 to TP-8) through the analysis of the [La] response. The third group was the sedentary control (SC, n = 9). The TP were classified in five intensity levels, three aerobic (A-1, A-2, A-3) and two anaerobic (AN-1, AN-2). Analysis of variance (ANOVA one-way, P<0.05) indicated significant differences in the [La] among the TP and among the five intensity levels. All TP of the A-2 and A-3 intensity levels differed from the A-1 and AN-1. The A-1 and AN-1 also differed among them. These findings demonstrate that the TP were classified properly at different levels of aerobic and anaerobic intensities, as based on the [La] response in a way similar to that of high performance swimming with humans. The results offer new perspectives for the study of exercise training in swimming rats at different levels intensity for performance or for health.

Improving accuracy and speed of optimum-path forest classifier using combination of disjoint training subsets

The Optimum-Path Forest (OPF) classifier is a recent and promising method for pattern recognition, with a fast training algorithm and good accuracy results. Therefore, the investigation of a combining method for this kind of classifier can be important for many applications. In this paper we report a fast method to combine OPF-based classifiers trained with disjoint training subsets. Given a fixed number of subsets, the algorithm chooses random samples, without replacement, from the original training set. Each subset accuracy is improved by a learning procedure. The final decision is given by majority vote. Experiments with simulated and real data sets showed that the proposed combining method is more efficient and effective than naive approach provided some conditions. It was also showed that OPF training step runs faster for a series of small subsets than for the whole training set. The combining scheme was also designed to support parallel or distributed processing, speeding up the procedure even more. © 2011 Springer-Verlag.

Is it possible to make pixel-based radar image classification user-friendly?

In this paper we would like to shed light the problem of efficiency and effectiveness of image classification in large datasets. As the amount of data to be processed and further classified has increased in the last years, there is a need for faster and more precise pattern recognition algorithms in order to perform online and offline training and classification procedures. We deal here with the problem of moist area classification in radar image in a fast manner. Experimental results using Optimum-Path Forest and its training set pruning algorithm also provided and discussed. © 2011 IEEE.

Improving Parkinson's disease identification through evolutionary-based feature selection

Parkinson's disease (PD) automatic identification has been actively pursued over several works in the literature. In this paper, we deal with this problem by applying evolutionary-based techniques in order to find the subset of features that maximize the accuracy of the Optimum-Path Forest (OPF) classifier. The reason for the choice of this classifier relies on its fast training phase, given that each possible solution to be optimized is guided by the OPF accuracy. We also show results that improved other ones recently obtained in the context of PD automatic identification. © 2011 IEEE.

Speeding up optimum-path forest training by path-cost propagation

In this paper we present an optimization of the Optimum-Path Forest classifier training procedure, which is based on a theoretical relationship between minimum spanning forest and optimum-path forest for a specific path-cost function. Experiments on public datasets have shown that the proposed approach can obtain similar accuracy to the traditional one but with faster data training. © 2012 ICPR Org Committee.