Papers by Cleber Zanchettin
On the Existence of a Threshold in Class Imbalance Problems
2015 IEEE International Conference on Systems, Man, and Cybernetics, 2015
A Voronoi Diagram Based Classifier for Multiclass Imbalanced Data Sets
2016 5th Brazilian Conference on Intelligent Systems (BRACIS), 2016
Enhancing batch normalized convolutional networks using displaced rectifier linear units: A systematic comparative study
Expert Systems with Applications
A Voronoi Diagram Based Classifier for Multiclass Imbalanced Data Sets
2016 5th Brazilian Conference on Intelligent Systems (BRACIS), 2016
Building Ensembles with Classifier Selection Using Self-Organizing Maps
2016 5th Brazilian Conference on Intelligent Systems (BRACIS), 2016
2008 Ieee International Joint Conference on Neural Networks, Jun 1, 2008
This paper investigates the problem of feature subset selection as part of a methodology that int... more This paper investigates the problem of feature subset selection as part of a methodology that integrates heuristic tabu search, simulated annealing, genetic algorithms and backpropagation. This technique combines both global and local search strategies for the simultaneous optimization of the number of connections and connection values of Multi-Layer Perceptron neural networks. We compare the performance of the proposed method for feature subset selection to five classical feature selection methods in three different classification problems.
This work investigates the use of Hybrid Intelligent Systems in the pattern recognition system of... more This work investigates the use of Hybrid Intelligent Systems in the pattern recognition system of an artificial nose. The connectionist approaches Multi-Layer Perceptron and Time Delay Neural Networks; and the hybrid approaches Feature-weighted Detector and Evolving Neural Fuzzy Networks were investigated. A wavelet filter as preprocessing method of odors signals is evaluated. The signals generated by an artificial nose, composed by an array of conducting polymer sensors, exposed to two different odor databases.
Fifth International Conference on Hybrid Intelligent Systems, Dec 6, 2005
Interest in hybrid methods that combine artificial neural networks and fuzzy inference systems ha... more Interest in hybrid methods that combine artificial neural networks and fuzzy inference systems has grown in recent years. These systems are robust solutions that search for representations of domain knowledge, reasoning on uncertainty, automatic learning and adaptation. However, the design and definition of the parameter effectiveness of such systems is still a hard task. In the present work, we perform a statistical analysis to verify interactions and interrelations between parameters in the design of neuro-fuzzy systems. The analysis is carried out using a powerful statistical tool, namely, Design of Experiments (DOE), in two neuro-fuzzy models-Adaptive Neuro Fuzzy Inference System (ANFIS) and Evolving Fuzzy Neural Networks (EFuNN). The results show that, for ANFIS, input MFs number and output MFs shape are usually the factors with the largest influence on the system's RMSE. For EFFuNN, the MF shape and the interaction between MF shape and number usually have the largest effect size.
International Journal of Hybrid Intelligent Systems, 2011
His, 2003
This work presents an application of a neuro-fuzzy model as a pattern recognition system for an a... more This work presents an application of a neuro-fuzzy model as a pattern recognition system for an artificial nose. The proposed model seeks to select important features among given plausible features while maintaining maximum recognition rate. The knowledge acquired by the network can be described as a set of interpretable rules. The results of neuro-fuzzy model are compared with two other widely used models of odor pattern classification, the Multi-Layer Perceptron (MLP) neural network and the Time Delay Neural Network (TDNN), in the analysis of signals generated by eight conducting polymer sensors exposed to gases derived from the petroliferous industry.
QRNN: q-Generalized Random Neural Network
IEEE Transactions on Neural Networks and Learning Systems, 2016
Artificial neural networks (ANNs) are widely used in applications with complex decision boundarie... more Artificial neural networks (ANNs) are widely used in applications with complex decision boundaries. A large number of activation functions have been proposed in the literature to achieve better representations of the observed data. However, only a few works employ Tsallis statistics, which has successfully been applied to various other fields. This paper presents a random neural network (RNN) with q-Gaussian activation functions [q-generalized RNN (QRNN)] based on Tsallis statistics. The proposed method employs an additional parameter q (called the entropic index) which reflects the degree of nonextensivity. This approach has the flexibility to model complex decision boundaries of different shapes by varying the entropic index. We conduct numerical experiments to analyze the efficiency of QRNN compared with RNNs and several other classical methods. Statistical tests (Wilcoxon and Friedman) are used to validate our results and show that the QRNN performs significantly better than RNNs with different activation functions. In addition, we find that QRNN outperforms many of the compared classical methods, with the exception of support vector machines, in which case it still exhibits a substantial advantage in terms of implementation simplicity and speed.
On the Existence of a Threshold in Class Imbalance Problems
2015 IEEE International Conference on Systems, Man, and Cybernetics, 2015
Extreme Learning Machine for Real Time Recognition of Brazilian Sign Language
2015 IEEE International Conference on Systems, Man, and Cybernetics, 2015
Face recognition based on global and local features
Proceedings of the 29th Annual Acm Symposium, Mar 24, 2014
Design and Application of Hybrid Intelligent Systems, 2003
This work presents results of the use of a wavelet filter for noise reduction and data compressio... more This work presents results of the use of a wavelet filter for noise reduction and data compression of signals generated by artificial nose sensors. To verify the performance of the wavelet analysis in the treatment of odor patterns, we compare two widely used artificial nose classifiers, multi-layer perceptron neural network and time delay neural network in the analysis of signals generated by eight conducting polymer sensors exposed to gases derived from the petroliferous industry.
Um Modelo Híbrido MLP-SVM para Reconhecimento de Caracteres Manuscritos Cursivos
This paper presents a hybrid MLP-SVM method for cursive characters recognition. Specialized Suppo... more This paper presents a hybrid MLP-SVM method for cursive characters recognition. Specialized Support Vector Machines (SVMs) are introduced to significantly improve the performance of Multilayer Perceptron (MLP) in the local areas around the surfaces of separation between each pair of characters in the space of input patterns. This hybrid architecture is based on the observation that when using MLPs in the task of handwritten characters recognition, the correct class is almost always one of the two maximum outputs of the MLP. The second observation is that most of the errors consist of pairs of classes in which the characters have similarities (e.g. (U, V), (m, n), (O, Q), among others). Specialized local SVMs are introduced to detect the correct class among these two classification hypotheses. The hybrid MLP-SVM recognizer showed improvement, significant, in performance in terms of recognition rate compared with an MLP for a task of character recognition. Resumo. Este artigo apresent...
In this paper a global and local optimization method is presented. This method is based on the in... more In this paper a global and local optimization method is presented. This method is based on the integration of the heuristic Simulated Annealing, Tabu Search, Genetic Algorithms and Backpropagation. The performance of the method is investigated in the optimization of Multi-layer Perceptron artificial neural network architecture and weights. The heuristics perform the search in a constructive way and based on the pruning of irrelevant connections among the network nodes. Experiments demonstrated that the method can also be used for relevant feature selection. Experiments are performed with four classification and one prediction datasets. 2 THE PROPOSED METHOD The proposed method (GaTSa) is based on the integration of the heuristic Simulated Annealing, Tabu Search, Genetic Algorithms and Backpropagation. The performance of the method is investigated in the simultaneous optimization of MLP architecture and weights. The pseudo-code of the proposed method and more implementation details are presented in (XXXXX, 2006). The next subsections presenting some important method implementation details for the paper.
A neural architecture to identify courtesy amount delimiters
The 2006 IEEE International Joint Conference on Neural Network Proceedings, 2006
Page 1. A neural architecture to identify courtesy amount delimiters1 Cleber Zanchettin, George D... more Page 1. A neural architecture to identify courtesy amount delimiters1 Cleber Zanchettin, George DC Cavalcanti, Rodrigo C. Dória, Eduardo FA Silva, Juliano CB Rabelo and Byron LD Bezerra AbstractThis paper deals with automatic recognition of real bank checks. ...
A hybrid intelligent system clonart for short and mid-term forecasting for the Brazilian Energy Distribution System
2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence), 2008
The present work describes an application of Clonart (Clonal Adaptive Resonance Theory) for forec... more The present work describes an application of Clonart (Clonal Adaptive Resonance Theory) for forecasting of amount of precipitation for the Brazilian Energy Distribution System. The effectiveness of the Brazilian electricity system directly depends on the difference between hydroelectric energy production and consumer use. Production depends upon the volume of water stored in the reservoirs. A forecasting system for the amount
Uploads
Papers by Cleber Zanchettin