Recent advances in Language Models have led to increasingly accurate models with enhanced capabilities to analyze and understand new datasets. This paper explores the use of open-source language models to extract judicial decisions from Portuguese consumer law texts. Our focus lies on the Named Entity Recognition (NER) task to identify and extract key information from documents. To this end, we evaluate models such as BERT, Meta’s Llama 3.1, and Google’s Gemma 2. The datasets used in this work were annotated and curated by domain-specialized lawyers. Experimental results demonstrate the effectiveness of applying Natural Language Processing (NLP) techniques along with transformer-based models for the NER task in Portuguese legal documents. We achieved results with higher accuracy with BERT models (F1-scores ranging from 0.74 to 0.96) for exact entity extraction, while LLMs showed promising results in zero-shot scenarios (F1-score of 0.85 for Llama 3). This work contributes to automating legal tasks and facilitating access to legal information in the Portuguese legal system.

In recent years, Representation Learning (RepL) has experienced a surge, especially for cross-modal applications with mixed-type data. Most current cross-modal RepL approaches rely mainly on supervised deep learning models, with unsupervised models playing auxiliary roles. In this paper, we present a fully unsupervised approach to demonstrate the capabilities of such models for cross-modal RepL. Our method jointly learns representations into topologically coherent cross-modal heteroassociative mappings. We apply the proposed framework to a yet to be solved multi-output prediction problem in petroleum geoscience: generating a complete set of regular petrophysical well logs from acoustic borehole images in highly heterogeneous Brazilian pre-salt reservoirs.

Semantic segmentation can be applied to a wide range of applications. One of the most interesting is medical image analysis. Applying semantic segmentation techniques in that field has great potential to assist physicians in diagnosing and analysing medical scans for the patient's benefit. Traditionally, several techniques have been applied in order to perform semantic segmentation. However, with the development of Deep Learning methods, there was a paradigm shift. Deep learning techniques are able to achieve great results, comparable to human-level performance. In order to achieve state-of-the-art results, researchers have to lean on the task of designing novel deep neural network architectures. That process is very time-consuming and heavily relies on experience and expert knowledge. Neural architecture search is the process of automatising the search for new deep neural network architectures. Quantum-Inspired Neural Architecture Search is a neural architecture search algorithm that leverages the benefits of quantum-inspired computing to search for neural network architectures efficiently. In this work, we adapt this to search for semantic segmentation neural networks and apply it to medical image analysis. The Spleen and Prostate datasets from the Medical Segmentation Decathlon challenge were used. Results show that our work was able to find better-performing semantic segmentation architectures for both datasets:$0.9583\pm 0.006$in comparison to ResU-Net$0.9525\pm 0.008$for the spleen dataset, and$0.6887\pm 0.067$in comparison to$0.6529\pm 0.070$for the prostate dataset.

The deployment of Deep Learning models requires attention to certain aspects not typically considered during the training phase. One of these is identifying and labeling samples from unknown classes. This is the goal of out-of-distribution (OOD) detection, which enhances the robustness of models in open-world scenarios. There are numerous methods for addressing this problem, using different feature spaces to distinguish between in-distribution and OOD data. One such method is the Open Principal Component Score (OpenPCS), a technique developed for open-set semantic segmentation using intermediate features of a fully convolutional network. This article introduces an extension of OpenPCS for multi-class classification, called OpenPCS-Class. We evaluate our approach in image and text classification tasks using various benchmark datasets and OOD detection methods. We also assess the effect of intermediate layers and network architectures on the OOD detection task. Our method outperformed other methods by up to 6.2% in terms of AUROC and 91% in terms of FPR95.

The field of radio astronomy is witnessing a boom in the amount of data produced per day due to newly commissioned radio telescopes. One of the most crucial problems in this field is the automatic classification of extragalactic radio sources based on their morphologies. Most recent contributions in the field of morphological classification of extragalactic radio sources have proposed classifiers based on convolutional neural networks. Alternatively, this work proposes gradient boosting machine learning methods accompanied by principal component analysis as data-efficient alternatives to convolutional neural networks. Recent findings have shown the efficacy of gradient boosting methods in outperforming deep learning methods for classification problems with tabular data. The gradient boosting methods considered in this work are based on the XGBoost, LightGBM, and CatBoost implementations. This work also studies the effect of dataset size on classifier performance. A three-class classification problem is considered in this work based on the three main Fanaroff-Riley classes: class 0, class I, and class II, using radio sources from the Best-Heckman sample. All three proposed gradient boosting methods outperformed a state-of-the-art convolutional neural networks-based classifier using less than a quarter of the number of images, with CatBoost having the highest accuracy. This was mainly due to the superior accuracy of gradient boosting methods in classifying Fanaroff-Riley class II sources, with 3-4% higher recall.

The use of computational intelligence models for multi-step time series forecasting tasks has presented satisfactory results in such a way that they are considered models with an excellent future for this type of problem. From the point of view of computational cost, the current alternatives combined with classical models are generating hybrid models that present even better results. Within the AutoML category, the optimization of hyperparameters and the selection of network topologies has become a challenge. Reservoir Computing, which is within the area of ​​Recurrent Neural Networks (RNN), proposes a particular model called Echo State Networks. which has been tested in different applications with excellent results; however, the difficulty in specifying the hyperparameters has been the subject of continuous study given the random nature of the set of neurons called Reservoir. Based on the Separation Ratio Graph (SRG) model for performance analysis, this paper proposes a new model, called Echo State Network - Genetic Algorithm - Separation Ratio Graph (ESN-GA-SRG), which optimizes network hyperparameters and at the same time selects the best topology for the Reservoir using the SRG coefficient, to find the reservoir that offers the most suitable dynamic behavior. The performance of this new model is evaluated on forecasting two sets of time series benchmarks with different characteristics of sampling periodicity, skewness, and stationarity. The results obtained show that the ESN-GA-SRG model was superior in predicting these time series in most cases, with statistical significance, when compared to other models that have been presented for this type of problem in the literature.

Covid-19 vaccine hesitancy and acceptance delay is an unprecedented challenge for concerned authorities. Existing studies lack the investigation about public vaccination acceptance, specifically for children. In this study, we surveyed the adult population in the UK to determine the diversity in public perception and acceptance of Covid-19 vaccination specifically for the children, among different sociodemographic groups. Statistical results and intelligent clustering outcomes indicate significant relationships between sociodemographic diversity and vaccination acceptance for children and their families. Acceptability for children is significantly dependent on ethnicity$(\mathrm{p}=3.7\mathrm{e}-05)$, age group, and gender, where only 47% of participants show willingness towards children's vaccination. Primary dataset in this study, along with the experimental outcomes, might be useful for public awareness and policy makers towards better preparation for future epidemics as well as working globally to combat the ongoing Covid-19 variations while running effective vaccination campaigns in the identified sociodemographic groups.

User eXperience (UX) has been used to achieve improvements in digital information systems based on how people perceive them. In particular, this paper establishes a framework that employs methods for eye and mouse tracking, keyboard input, self-assessment questionnaire and artificial intelligence algorithms to evaluate user experience and categorize users in terms of performance profiles. The results obtained with this framework are artifacts that can be used to support customizations of the User Interface (UI) on the websites. Moreover, the established framework is generic and flexible and can be applied to any information system, such as the case study shown in the website of the Federal Revenue of Brazil (RFB). The main objectives of this paper are as follows: (i) to set out a powerful UX framework based on three tracking techniques – the AIT2-UX; (ii) to provide the T2-UXT to collect, collate, process and visualize data obtained from users’ interactions (iii) to use and compare machine learning algorithms with the classification of user performance profiles; (iv) to use the artifacts generated by the framework to manually customize the UI with the website.

Advances in multi-omic data collection technologies contributed to the availability of molecular information in different levels, which progressively increased redundancy, complexity, and volume of datasets. This study developed a methodology for clinical and multi-omic data integration, with the aim of improving multi-omic data selection regarding cancer molecular subtypes identification. Tumor characterization into molecular subtypes can promote a more personalized and assertive medical practice. As a case study, methylation, miRNA expression, and mRNA expression data from colorectal cancer (CRC) samples were acquired from The Cancer Genome Atlas (TCGA) project portal. CRC has the fourth highest incidence in the world population and the third highest in Brazil. The datasets were first pre-processed by applying feature selection techniques to reduce the dimensionality. Then, Random Forest classification model was used to identify the molecular subtypes of CRC, based on the significantly reduced amount of features. The association of clinical and omics data showed promising results concerning phenotype prediction.

Forecasting stock prices in the market its known to be an extremely difficult task, where even the predictability of the series itself is a controversial matter. The present study investigates the existence of periods within the series more suitable for prediction, and whether the identification and exploitation of those periods could be learned from data. In order to do that, the Predictability Crawler (P-Craw) framework is proposed. The technique uses optimizations routines such as the Particle Swarm optimization (PSO) or Genetic Algorithms (GA) to select subsets of historical data where statistical learning algorithms can be more efficiently trained. When tested against simulated data, The P-Craw is able to reliably identify the optimal subsets in scenarios ranging from 40% to 100% of predictable samples in the data. To access if the framework brings any improvement when used in a real world scenario, it is tested in a dataset containing intraday data from the Brazilian stocks exchange (BOVESPA). When benchmarked against training with all the samples for the series in the BOVESPA dataset the use of the framework is able to significantly raise the Correct Directional Changes (CDC) of the trained models while reducing the Mean Absolute Error (MAE) in up to 19%.

In order to develop an offshore maintenance schedule support system, this work presents a new model for constrained combinatorial problems: CPSO+. This model is a combination of two previous models: the PSO+, which presented good results in problems with nonlinear constrains; and the CPSO, which is an adaptation of PSO for application in combinatorial problems. The proposed model has been adapted to solve the complex problem of defining the best sequence of offshore maintenance activities and allocated staff to maximize service provider profitability within three months, while respecting all service completion time constraints and specific offshore work constraints. To evaluate the performance of this new model in solving the proposed problem, two CPSO+ variants were evaluated against the original CPSO model, in six proposed simulation cases. The results of the simulations indicate that the proposed CPSO+ model with reduced initialization variation outperforms other evaluated models in execution time and solution quality to the given problem.

Quantum-inspired Evolutionary Algorithms (QiEAs) have demonstrated to be very effective in several applications. In particular, employing this algorithm for feature selection as a wrapper technique in Brain-Computer Interfaces applications was recently proposed with great results. Moreover, the training time of the model was decreased while maintaining a high classification accuracy, both essential conditions for a successful BCI. The drawback of this model was the sensitiveness to changes in the direction and magnitude of the rotation angle, which can produce adverse effects in both performance and convergence time. Chaotic systems and Evolutionary algorithms, when combined, can enhance the convergence rate and speed of the evolutionary process, incrementing the capacity of reaching the global optima. In this paper we explore the effects of adding ergodicity to a QiEA by the employment of chaotic maps in two operators: chaotic uniform crossover and chaotic quantum update gate. To validate the proposed approach, six commonly used chaotic maps are tested with data of Motor Imagery (MI) Electroencephalography (EEG) of right and left hand movement. The results of these experiments are compared with the ones of a QiEA and a classical Genetic Algorithm (GA). In the proposed model, Wavelet Packet Decomposition is employed as the time-frequency analysis to characterize the signal, whereas a Multilayer Perceptron Neural Network is used as a classifier. The results demonstrated that Chaotic QiEAs can significantly improve the convergence time of the model with only a small loss in the final accuracy.

Abstract This work presents a new neuro-evolutionary model, called NEVE (Neuroevolutionary Ensemble), based on an ensemble of Multi-Layer Perceptron (MLP) neural networks for learning in nonstationary environments. NEVE makes use of quantum-inspired evolutionary models to automatically configure the ensemble members and combine their output. The quantum-inspired evolutionary models identify the most appropriate topology for each MLP network, select the most relevant input variables, determine the neural network weights and calculate the voting weight of each ensemble member. Four different approaches of NEVE are developed, varying the mechanism for detecting and treating concepts drifts, including proactive drift detection approaches. The proposed models were evaluated in real and artificial datasets, comparing the results obtained with other consolidated models in the literature. The results show that the accuracy of NEVE is higher in most cases and the best configurations are obtained using some mechanism for drift detection. These results reinforce that the neuroevolutionary ensemble approach is a robust choice for situations in which the datasets are subject to sudden changes in behaviour.

In recent years, the use of mobile robots in agriculture has increased significantly because of their capability to carry out agricultural tasks in a safe and efficient manner, with limited or without human intervention. Crop monitoring and inspection have become an important part of precision agriculture, supporting farmers in the management of insect pests, weeds and diseases in order to reduce costs and losses. In this work, we present the design and development of an autonomous mobile robot, conceived to perform routine monitoring and inspection tasks on soy and cotton crops. The robot design is similar to a differential-drive vehicle due to its simplicity of construction, modeling and control. The autonomous navigation is successfully carried out via the use of odometry and cameras properly attached to the robot structure. Preliminary field tests with the prototype operating on a cotton farm are presented to show the performance and feasibility of the electro-mechanical design for navigation in row crops.

In this work, we address the modelling and control design of a wheeled mobile robot capable of performing autonomous navigation tasks in agricultural fields. The methodology is based on the kinematics approach due to its wellknown ability to ensure satisfactory performance when the robot motions are carried out with low velocities and slow accelerations. Two control strategies are used for stabilization and trajectory tracking purposes: the first scheme is based on a fuzzy logic algorithm and considers the regulation problem of the robot position in Cartesian space; the second scheme is based on a visual servoing algorithm and considers the tracking problem of a straight line by using a constant linear velocity in Cartesian space and the position error in image space. Simulation results are presented to illustrate the effectiveness and feasibility of the fuzzy logic approach. Experimental tests with an agricultural mobile robot are carried out to verify and validate the visual servoing approach for cotton crop monitoring.

Agriculture is considered the economic basis of countries around the globe, and the development of new technologies contributes to the harvesting efficiency. Autonomous vehicles are used in farms for seeding, harvesting and tasks like pesticide application. However, one of the main issues of any plantation is insect pest and disease identification, essential for pest control and maintenance of healthy plants. This work presents and compares three methods for insect pest identification using computer vision: Deep Convolutional Neural Network (DCNN), as a baseline; Hierarchical Deep Convolutional Neural Network (HD-CNN), in order to improve prediction of similar classes; and Pixel-wise Semantic Segmentation Network (SegNet). They were tested for two kinds of culture, soybean and cotton. SegNet outperformed both approaches by a wide margin: the methods had respective accuracies of 70.14% DCNN, 74.70% HD-CNN and 93.30% SegNet.

Fuzzy Inference Systems (SIF) for classification are machine learning models that employ linguistic rules to describe real problems in an interpretable way. However, when dealing with high dimensional input spaces, these systems tend to suffer from scalability problems and computational cost. This work presents the Selected Ensemble of Fuzzy Inference Systems (SENFIS), an automatic SIF model based on previously developed algorithms (AutoFIS-Class and RandomFIS) and composed of ensemble and subsampling strategies to deal with big datasets. Its performance is compared to those of its predecessors and of others similar fuzzy systems, making use of normal and large benchmarks databases for classification. In terms of accuracy, SENFIS performs as well as its predecessors, but at a lower computational cost. In addition, the resulting rule bases are smaller, especially for high dimensional problems.

Currently, Health Units in a large number of countries in the world present service demand that exceeds their real capacities. This problem causes the inevitable emergence of long waiting lists. The optimization of such waiting list is very challenging, due to the large number of resources that must be considered. This paper proposes a new model, based on a quantum-inspired evolutionary algorithm, to optimize the scheduling of for elective surgical procedures. The Quantum-Inspired Evolutionary Algorithm for Healthcare (QIEA-H) model, aims not only to designate the necessary resources to the patients in order to achieve the successful completion of the chirurgical procedure, but also to reduce the total time used to perform all surgeries and the number of surgeries out of term. For the validation of the proposed model, a waiting list of 2000 surgeries was created artificially and using a simulation tool also developed in this work. The model achieved a reduction in the time of all surgeries of up to 16.25% and the number of surgeries out of date of up to 13.04%.

Degradation of speech signal due to adverse conditions is the major challenge for automatic speech recognition (ASR) systems. This paper introduces a novel approach to estimate an Ideal Neighbourhood Mask (INM) for speech segregation based on deep neural networks estimator. The method described here is based on the local binary patterns (LBP) technique often used in digital image processing. Ideal Neighbourhood Mask will indicate which time-frequency (T-F) units of the noisy speech are canceled. The performance assessment of the proposed application in conjunction with the traditional mask techniques, i.e., Ideal Binary Mask (IBM) and Ideal Ratio Mask (IRM), are carried out under various environments regarding the objective speech quality measures. The recognition experiments including results in the AURORA IV framework indicate that the proposed scheme, when applied in adverse environments yield significantly better performance than the conventional techniques.

Deep neural networks have gained attention in the last decade as significant progress has been made in a variety of tasks thanks to these new architectures. Most of the time, hand-designed networks are responsible for this incredible success. However, this engineering process demands considerable time and expert knowledge, which leads to an increasing interest in automating the design of deep architectures. Several new algorithms have been proposed to address the neural architecture search problem, but many of them require significant computational resources. Quantum-inspired evolutionary algorithms (QIEA) have their roots on quantum computing principles and present promising results in respect to faster convergence. In this work, we propose Q-NAS (Quantum-inspired Neural Architecture Search): a quantum-inspired algorithm to search for deep neural architectures by assembling substructures and optimizing some numerical hyperparameters. We present the first results applying Q-NAS on the CIFAR-10 dataset using only 20 K80 GPUs for about 50 hours. The obtained networks are relatively small (less than 20 layers) compared to other state-of-the-art models and achieve promising accuracies with considerably less computational cost than other NAS algorithms.

The occurrence of a disaster brings about damages, destruction, ecological disruption, loss of human life, human suffering, deterioration of health and health service of sufficient magnitude to require external assistance, demanding the mobilization and deployment of emergency rescue units within the affected area, in order to reduce casualties and economic losses. The scheduling of those units is one of the key issues in the emergency response phase and can be seen as a generalization of the unrelated parallel machine scheduling problem with sequence and machine dependent setup. The objective is to minimize the total weighted completion time of the incidents to be attended, where the weight correspond to its severity level. We propose a biased random-key genetic algorithm to tackle this problem, considering fuzzy required processing times for the incidents, and compare the solutions with those generated by a constructive heuristic, from the literature, developed to deal with this problem. Our results show that the genetic algorithm's solutions are 2.17% better than those obtained with the constructive heuristic when applied to instances with up to 40 incidents and 40 rescue units.

Quantum-Inspired Evolutionary Algorithms (QIEA) represent an efficient alternative to the traditional genetic algorithms, being capable of finding good solutions with smaller populations. Echo State Networks (ESNs) are a simple and efficient implementation of the Reservoir Computing framework. The use of this kind of networks in system identification is advantageous due to its intrinsic dynamic behavior and fast training procedure. However, ESNs have global parameters that should be tuned in order to improve their performance in a determined task. Besides, the random generation of the reservoir weights of these networks may not be ideal in terms of performance. Thus, this work presents a method that automatically defines an ESN for system identification problems by using a real coded QIEA (QIEA-R) in a two-phase optimization procedure. The QIEA-R firstly searches for the best global parameters of an ESN; then, on a second stage, optimizes some of its reservoir weights. In two benchmark problems for system identification, the proposed method overcame the performance of a randomly generated ESN with the same global parameters and has presented comparable and, in most cases, better accuracy results in comparison to some methods which were applied to the same datasets.

In Brain-Computer Interfaces, one of the most relevant tasks is the selection of a subset of features that efficiently describes the EEG signal, excluding redundant and irrelevant features. This procedure reduces the dimensionality of the dataset (avoiding the dimensionality curse) and improves the classification accuracy of the system. One of the most successful models applied for this task is the use of an Evolutionary Algorithm in a wrapper approach. These models produce excellent results but present the drawback of a considerable high processing time, a critical limitation for its application on real Brain-Computer Interfaces (BCI) systems. Quantum-inspired Evolutionary Algorithms can be an alternative wrapper approach for the feature selection task, given that they outperform classical Evolutionary Algorithms in the exploration and exploitation of the search space, obtaining the global solution much faster. These algorithm employs concepts and principles from the Quantum Mechanics to probabilistically describe a set of different states between the classical logic states 0 and 1. In this paper, a Quantum-inspired Evolutionary Algorithm is developed and tested over three different subjects from publicly available datasets. In the proposed model, Wavelet Packet Decomposition is employed to analyze the time-frequency characteristics of the signals, and a Multilayer Perceptron Neural Network is employed as a classifier.

Gas turbines are the most common engine used in the majority of commercial aircraft. Due to its criticality, to detect and classify faults in a gas turbine is extremely important. In this work, a type-1 and singleton fuzzy logic system trained by steepest descent method is used for detecting and classifying gas turbine faults. The data set was obtained through simulations on the software Propulsion Diagnostic Method Evaluation Strategy created by the National Aeronautics and Space Administration. Results are compared to those obtained with a type-1 fuzzy classifier with rule extraction by Wang and Mendel method. Analysis of results shows the effectiveness of the proposed model. When compared to the Wang and Mendel fuzzy classifier, it requires fewer rules to achieve a better performance.

A switch machine is an electromechanical device that allows railway trains to be guided from one track to another. Among all possible faults that can occur in a switch machine, the three mains ones are: lack of lubrication, lack of adjustment and malfunction of a component. Aiming to classify these faults, an important contribution of this work is to address the height type-reduction and interval singleton type-2 fuzzy logic system derivatives. The computational simulations are performed with real data set provided by a Brazilian company of the railway sector. The obtained results are compared with other models reported in the literature (Bayes theory, multilayer perceptron neural network and type-1 fuzzy logic system), demonstrating the effectiveness of the proposed classifier and revealing that the proposal is able to properly handle with uncertainties associated with the measurements and with the data that are used to tune the parameters of the model. In addition, the convergence speed and performance analysis show that the proposed interval singleton type-2 fuzzy logic system is attractive for classifying faults in a switch machine.

This article discusses the practical aspects of using neural networks in the identification of nonlinear industrial plants through the use of closed-loop operation data, with the purpose of creating a computational model that facilitates the development and tuning of intelligent control algorithms. In this paper, specifically, an oil separation plant is identified and the resulting model is employed to develop a fuzzy logic controller.

We present Random Forest, Support Vector Machine and Feedforward Neural Network models to classify 2519 variable star light curves. These light curves are generated from a reduction of non-survey optimized observational images gathered by wide-field cameras mounted on the Liverpool Telescope. We extract 16 features found to be highly informative in previous studies and achieve an area under the curve of 0.8495 using a feedforward neural network with 50 hidden neurons trained with stratified 10-fold cross-validation with 3 repeats. We propose using an automated visual feature extraction technique by transforming bin-averaged phase-folded light curves into image based representations. This eliminates much of the noise and the missing phase data, due to sampling defects, should have a less destructive effect on these shape features as they still remain at least partially present. There is also no need for feature engineering as the learning algorithms can learn shape features directly from the light curves. We produced a set of scaled images based on a threshold of data points in each pixel. Training on the same feedforward network, we achieve an area under the curve of 0.6348. By introducing the Period and Amplitude as features into this dataset therefore giving meaning to the dimensions of the image we show this improves to 0.7952. Our current models lack translational-invariance and the method may be better suited to specific sub-classification problems common in the variable object hierarchical multi-class problem.

Brain Computer Interfaces allow the interaction between a person and their environment using signals extracted directly from the brain. One of the most common non-invasive methods of brain signal acquisition is the electroencephalography (EEG). An EEG based BCI system generally involves four steps: preprocessing, feature extraction, feature selection, and classification. In order to design a real applicable BCI system, it is important to provide: good classification performance, adequate computational cost, robustness to variations of the signal between trials and between subjects, and a classifiers model that cope with highly dimensional data. Ensemble of classifiers is a learning model that, with a proper design, can satisfy those conditions, which make them a good match for a BCI application. In this paper, some ensemble of classifiers designs are evaluated and compared with other BCI approaches on three different subjects. In the proposed model, Genetic Algorithm is employed as feature selection method and Wavelet Packet Decomposition as preprocessing procedure. Four fusion methods were applied in the ensembles design, including: Majority Voting, Weighted Majority Voting, Genetic Algorithm for classifier selection and Genetic Algorithm to compute the weights for the Weighted Majority fusion method. The best results were obtained with the Weighted Majority Voting fusion method based on Genetic Algorithm.

The function of each protein in the body is determined by its 3D structure, which can be predicted by computational methods. These methods generate an exceptional quantity of candidate models (decoys). similarity and machine learning methods are used to assess their quality. When measuring the distance from the decoy to its native structure (RMSD, TM-Score, Z-Score), similarity methods may be applied. On the other hand, machine learning methods use a subset of structural and physicochemical features to assess the quality of these candidate models. In the preprocessing step, a subset of these features is selected by hand to be used in the machine learning process. The model proposed in this work considers different sets of features simultaneously, and automatically selects, via an evolutionary model, the optimal subset to be used in the machine learning method. The proposed model also provides the relative feature importance related to the quality of the decoy model. The new model, named Score Wrapper to Feature Importance Calculation (SWtoFIC), also calculates the quality of the decoy model. These characteristics make this model an important tool to assess the decoy quality and to better understand the influence of different types of features in the decoy quality determination.

This paper focuses on the classification of faults in an electromechanical switch machine, which is an equipment used for handling railroad switches. In this paper, we introduce the use of Set-Membership concept, derived from the adaptive filter theory, into the training procedure of type-1 and singleton/non-singleton fuzzy logic systems, in order to reduce computational complexity and to increase convergence speed. We also present different criteria for using along with Set-Membership. Furthermore, we discuss the usefulness of delta rule delta, local Lipschitz estimation, variable step size, and variable step size adaptive techniques to yield additional improvement in terms of computational complexity reduction and convergence speed. Based on data set provided by a Brazilian railway company, which covers the four possible faults in a switch machine, we present performance analysis in terms of classification ratio, convergence speed, and computational complexity reduction. The reported results show that the proposed models result in improved convergence speed, slightly higher classification ratio, and remarkable computation complexity reduction when we limit the number of epochs for training, which may be required due to real-time constraint or low computational resource availability.

This paper proposes a new approach to Particle Swarm Optimization (PSO) to solve nonlinear problems with linear and nonlinear constraints. A crossover operator and a new particle updating method, named Footholds Concept, were developed to guarantee fully feasible solutions and better search-space coverage, respectively. In addition, a novel swarm initialization heuristic is applied to benchmarks with equality constraints. The algorithm has been tested on 13 common benchmark functions. Experimental results show that it is very competitive as it increases PSO efficiency and improves convergence speed.

It can be argued that in order to produce a sub-stantial improvement in multi-objective estimation of distribution algorithms it is necessary to focus on a particular group of issues, in particular, on the weaknesses derived from multi-objective fitness assignment and selection methods, the incorrect treatment of relevant but isolated (precursor) individuals; the loss of population diversity, and the use of ‘general purpose’ modeling algorithms without taking note of the particular requirements of the task. In this work we introduce the S-Metric Selection Estimation of Distribution Algorithm based on Multivariate Extension of Copulas (SMS-EDA-MEC). SMS-EDA-MEC was devised with the intention of dealing with those issues in mind. It builds the population model relying on the comprehensive Clayton's copula and incorporates methods for automatic population restarting and for priming precursor individuals. The experimental studies presented show that SMS-EDA-MEC yields better results than current and ‘traditional’ approaches.

A time series is the most commonly used representation for the evolution of a given variable over time. In a time series forecasting problem, a model aims at predicting the series' future values, assuming that all information needed to do so is contained in the series' past behavior. Since the phenomena described by the time series does not always exist in isolation, it is possible to enhance the model with historical data from other related time series. The structure formed by several different time series occurring in parallel, each featuring the same interval and dimension, is called a multivariate time series. This paper presents a methodology for the generation of a Fuzzy Inference System (FIS) for multivariate time series forecasting from historical data, aiming at good performance in both forecasting accuracy and rule base interpretability - in order to extract knowledge about the relationship between the modeled time series. Several aspects related to the operation and construction of such a FIS are investigated regarding complexity and semantic clarity. The model is evaluated by applying it to multivariate time series obtained from the complete M3 competition database and by comparing it to other methods in terms of accuracy. In addition knowledge extraction possibilities from the resulting rule base are explored.

This work presents a novel genetic fuzzy system for forecasting, called Genetic Programming Fuzzy Inference System for Forecasting problems (GPFIS-Forecast), which generates an interpretable fuzzy rule base by using Multi-Gene Genetic Programming to define the premises terms of fuzzy rules. The main differences between GPFIS-Forecast and other genetic fuzzy systems lie in its fuzzy inference process, because it: (i) enables premises to be include negation, t-conorm and linguistic hedge operators; (ii) applies methods to define a consequent term more compatible with a given premise; and (iii) makes use of aggregation operators to weigh fuzzy rules in accordance with their influence on the problem. GPFIS-Forecast has been tested in the NN3 Competition, in order to evaluate its performance in a benchmark problem. In this case, it has produced competitive results when compared to other forecasting approaches.

The brain of the fruit fly Drosophila Melanogaster is an attractive system for studying the logic underlying neural circuits because it implements a rich behavior repertoire with a number of neural components that is five orders of magnitude smaller than that of vertebrates. Analysis of the fly's connectome using a powerful toolkit of well-developed genetic techniques and advanced electrophysiological recording tools enables the fly's neural circuitry to be experimentally mapped into functional units, called Local Processing Units (LPU). Many tools are already available to enable neuroscientists to create an accurate model of the entire fly brain, but none of them provides a method to specify those circuits in a way that both biologists and engineers can work together. Also, the development of plausible LPU models requires the ability to specify and instantiate subcircuits without explicit reference to their constituent neurons and internal connections. To this end, we present a neural circuit specification language called CircuitML for construction of LPUs. CircuitML has been designed as an extension to NeuroML; it provides constructs for defining subcircuits that comprise neural primitives supported by NeuroML. Subcircuits are endowed with interface ports that enable connections to other subcircuits via neural connectivity patterns. We have used CircuitML to specify an LPU-based model of the fly olfactory system.

Most drift detection mechanisms designed for classification problems works in a post-event manner: after receiving the data set completely (patterns and class labels of the train and test set), they apply a sequence of procedures to identify some change in the class-conditional distribution - a concept drift. However, detecting changes after its occurrence can be in some situations harmful for the process under supervision. This paper proposes a pre-event approach for abrupt drift detection, called by A2D2. Briefly, this method is composed of three steps: (i) label the patterns from the test set, using an unsupervised method; (ii) compute some statistics from the train and test set, conditioned on the given class labels; and (iii) compare the train and test statistics using a multivariate hypothesis test. Also, it has been proposed a procedure for creating datasets with abrupt drift. This procedure was used in the sensivity analysis of A2D2, in order to understand the influence degree of each parameter on its final performance.

This work presents a Genetic Fuzzy Controller (GFC), called Genetic Programming Fuzzy Inference System for Control tasks (GPFIS-Control). It is based on Multi-Gene Genetic Programming, a variant of canonical Genetic Programming. The main characteristics and concepts of this approach are described, as well as its distinctions from other GFCs. Two benchmarks application of GPFIS-Control are considered: the Cart-Centering Problem and the Inverted Pendulum. In both cases results demonstrate the superiority and potentialities of GPFIS-Control in relation to other GFCs found in the literature.

One of the most important issues in oil \& gas industry is the lithological identification. Lithology is the macroscopic description of the physical characteristics of a rock. This work proposes a new methodology for lithological discrimination, using GPF-CLASS model (Genetic Programming for Fuzzy Classification) a Genetic Fuzzy System based on Multi-Gene Genetic Programming. The main advantage of our approach is the possibility to identify, through seismic patterns, the rock types in new regions without requiring opening wells. Thus, we seek for a reliable model that provides two flexibilities for the experts: evaluate the membership degree of a seismic pattern to the several rock types and the chance to analyze at linguistic level the model output. Therefore, the final tool must afford knowledge discovery and support to the decision maker. Also, we evaluate other 7 classification models (from statistics and computational intelligence), using a database from a well located in Brazilian coast. The results demonstrate the potentialities of GPF-CLASS model when comparing to other classifiers.

In our previous works [1, 2], we proposed NEVE, a model that uses a weighted ensemble of neural network classifiers for adaptive learning, trained by means of a quantum-inspired evolutionary algorithm (QIEA). We showed that the neuro-evolutionary classifiers were able to learn the dataset and to quickly respond to any drifts on the underlying data. Now, we are particularly interested on analyzing the influence of an unlimited ensemble, instead of the limited ensemble from NEVE. For that, we modified NEVE to work with unlimited ensembles, and we call this new algorithm NEVE++. To verity how the unlimited ensemble influences the results, we used four different datasets with concept drift in order to compare the accuracy of NEVE and NEVE++, using two other existing algorithms as reference.

This paper introduces two new hybrid models for clustering problems in which the input features and parameters of a spiking neural network (SNN) are optimized using evolutionary algorithms. We used two novel evolutionary approaches, the quantum-inspired evolutionary algorithm (QIEA) and the optimization by genetic programming (OGP) methods, to develop the quantum binary-real evolving SNN (QbrSNN) and the SNN optimized by genetic programming (SNN-OGP) neuro-evolutionary models, respectively. The proposed models are applied to 8 benchmark datasets, and a significantly higher clustering accuracy compared to a standard SNN without feature and parameter optimization is achieved with fewer iterations. When comparing QbrSNN and SNN-OGP, the former performed slightly better but at the expense of increased computational effort.

This paper presents the research and development of two hybrid neuro-fuzzy models for the hierarchical coordination of multiple intelligent agents. The main objective of the models is to have multiple agents interact intelligently with each other in complex systems. We developed two new models of coordination for intelligent multiagent systems, which integrates the Reinforcement Learning Hierarchical Neuro-Fuzzy model with two proposed coordination mechanisms: the MultiAgent Reinforcement Learning Hierarchical Neuro-Fuzzy with a market-driven coordination mechanism (MA-RL-HNFP-MD) and the MultiAgent Reinforcement Learning Hierarchical Neuro-Fuzzy with graph coordination (MA-RL-HNFP-CG). In order to evaluate the proposed models and verify the contribution of the proposed coordination mechanisms, two multiagent benchmark applications were developed: the pursuit game and the robot soccer simulation. The results obtained demonstrated that the proposed coordination mechanisms greatly improve the performance of the multiagent system when compared with other strategies.

This work presents a Genetic Fuzzy Classification System (GFCS) called Genetic Programming Fuzzy Classification System (GPF-CLASS). This model differs from the traditional approach of GFCS, which uses the metaheuristic as a way to learn “if-then” fuzzy rules. This classical approach needs several changes and constraints on the use of genetic operators, evaluation and selection, which depends primarily on the metaheuristic used. Genetic Programming makes this implementation costly and explores few of its characteristics and potentialities. The GPF-CLASS model seeks for a greater integration with the metaheuristic: Multi-Gene Genetic Programming (MGGP), exploring its potential of terminals selection (input features) and functional form and at the same time aims to provide the user with a comprehension of the classification solution. Tests with 22 benchmarks datasets for classification have been performed and, as well as statistical analysis and comparisons with others Genetic Fuzzy Systems proposed in the literature.

This work describes the use of a weighted ensemble of neural network classifiers for adaptive learning. We train the neural networks by means of a quantum-inspired evolutionary algorithm (QIEA). The QIEA is also used to determine the best weights for each classifier belonging to the ensemble when a new block of data arrives. We show that the neuroevolutionary classifiers are able to learn the data set and to quickly respond to any drifts on the underlying data. We also compare the results reached by our model with an existing algorithm, Learn++.NSE, in two different nonstationary scenarios.

This work describes the use of a weighted ensemble of neural network classifiers for adaptive learning. We train the neural networks by means of a quantum-inspired evolutionary algorithm (QIEA). The QIEA is also used to determine the best weights for each classifier belonging to the ensemble when a new block of data arrives. After running several simulations using two different datasets and performing two different analysis of the results, we show that the proposed algorithm, named neuro-evolutionary ensemble (NEVE), was able to learn the data set and to quickly respond to any drifts on the underlying data, indicating that our model can be a good alternative to address concept drift problems. We also compare the results obtained by our model with an existing algorithm, Learn++.NSE, in two different nonstationary scenarios.

This article proposes a new algorithm based on evolutionary computation and quantum computing. It attempts to resolve ordering combinatorial optimization problems, the most well known of which is the traveling salesman problem (TSP). Classic and quantum-inspired genetic algorithms based on binary representations have been previously used to solve combinatorial optimization problems. However, for ordering combinatorial optimization problems, order-based genetic algorithms are more adequate than those with binary representation, since a specialized crossover process can be employed in order to always generate feasible solutions. Traditional order-based genetic algorithms have already been applied to ordering combinatorial optimization problems but few quantum-inspired genetic algorithms have been proposed. The algorithm presented in this paper contributes to the quantum-inspired genetic approach to solve ordering combinatorial optimization problems. The performance of the proposed algorithm is compared with one order-based genetic algorithm using uniform crossover. In all cases considered, the results obtained by applying the proposed algorithm to the TSP were better, both in terms of processing times and in terms of the quality of the solutions obtained, than those obtained with order-based genetic algorithms.

Online retail stores face great challenges to recommend products due to the size and sparsity of the databases, as well as the variety of new users and items. As current techniques, based on collaborative filtering, address those issues with only partial success, the present paper proposes the use of a hybrid system of recommendation in online stores. This system makes use of collaborative filtering and of a fuzzy number model based on marketing concepts. Experimental results show that the proposed system presents great invariance to sparse databases, which is of great value for retail companies.

This paper presents the proposal of a generic model of stochastic process based on neural networks, called Neural Stochastic Process (NSP). The proposed model can be applied to problems involving phenomena of stochastic behavior and / or periodic features. Through the NSP's neural networks it is possible to capture the historical series' behavior of these phenomena without requiring any a priori information about the series, as well as to generate synthetic time series with the same probabilities as the historical series. The NSP was applied to the treatment of monthly inflows series and the results indicate that the generated synthetic series exhibit statistical characteristics similar to historical series.

Composite materials have changed the way of using polymers, as the strength was favored by the incorporation of fibers and particles. This new class of materials allowed a larger number of applications. The insertion of nanometric sized particles has enhanced the variation of properties with a smaller load of fillers. In this paper, we attempt to a better understanding of nanocomposites by using an artificial intelligence's technique, known as artificial neural networks. This technique allowed the modeling of Young's modulus of nanocomposites. A good approximation was obtained, as the correlation between the data and the response of the network was high, and the error percentage was low.

Since they were proposed as an optimization method, the evolutionary algorithms have been successfully used for solving complex problems in several areas such as, for example, the automatic design of electronic circuits and equipments, task planning and scheduling, software engineering and data mining, among many others. However, some problems are computationally intensive when it concerns the evaluation of solutions during the search process, making the optimization by evolutionary algorithms a slow process for situations where a quick response from the algorithm is desired (for instance, in online optimization problems). Several ways to overcome this problem, by speeding up convergence time, were proposed, including Cultural Algorithms and Coevolutionary Algorithms. However, these algorithms still have the need to evaluate many solutions on each step of the optimization process. In problems where this evaluation is computationally expensive, the optimization can take a prohibitive time to reach optimal solutions. This work presents an evolutionary algorithm for numerical optimization problems (Quantum-Inspired Evolutionary Algorithm for Problems based on Numerical Representation - QIEA-R), inspired in the concept of quantum superposition, which allows the optimization process to be carried on with a smaller number of evaluations. It extends previous works by presenting a broader range of tests and improvements on the algorithm. The results show the good performance of this algorithm in solving numerical problems.

This work proposes a new approach for history matching using Kernel PCA to adjust the reservoir permeability field obeying geostatistical constraint. Although there are several methodologies in literature for history matching, most of them don't take into account geostatistical restrictions. Besides, history matching is a problem of huge dimensionality. So, Kernel PCA was chosen due to its ability to compress and accurately reconstruct data in addition to being able to extract non-linear characteristics.

This article describes a novel framework for combining time series forecasts. It uses neural network regression models to estimate, at a given point in time, the linear weights (relevancies) of the available experts (forecasters) at that time. With those weights, the experts can be linearly combined to produce a single, potentially more accurate, forecast. This new weight generation framework was designed to be especially useful for multi-step-ahead forecasting.

The volume of energy loss that Brazilian electric utilities have to deal with has been ever increasing. The electricity concessionaries are suffering significant and increasing loss in the last years, due to theft, measurement errors and many other kinds of irregularities. Therefore, there is a great concern from those companies to identify the profile of irregular customers, in order to reduce the volume of such losses. This paper presents the proposal of an intelligent system, composed of two neural networks ensembles, which intends to increase the level of accuracy in the identification of irregularities among low tension consumers. The data used to test the proposed system are from Light S.A. Company, the Rio de Janeiro concessionary. The results obtained presented a significant increase in the identification of irregular customers when compared to the current methodology employed by the company.

This paper presents a novel interval type-2 fuzzy inference system with automatic learning for handling uncertainty, called the hierarchical type-2 neuro-fuzzy BSP model (T2-HNFB). This new model combines the paradigms of the type-2 fuzzy inference systems and neural networks with recursive partitioning techniques (BSP – Binary Space Partitioning). The model is able to automatically create and expand its own structure, to reduce limitations on the number of inputs and to extract fuzzy linguistic rules from a dataset, as well as to efficiently model and manipulate most of the types of uncertainty existing in real situations. In addition, it provides a confidence interval for its output, which constitutes important information for real applications. In this context, this model overcomes the limitations of the conventional type-2 and type-1 fuzzy inference systems. Experimental results show that the results provided by the T2-HNFB model are close to and in several cases better than the best results supplied by the other models used for comparison.

Power transformers are one of the most used and expensive equipments in many substations of electric energy. This fact justifies the application of predictive techniques of diagnosis, with the objective to minimize possible failures and to increase the trustworthiness of the system. Amongst these techniques, one of the most distinguished are the analysis of gases dissolved in the oil (gaseous chromatography) and the physical-chemical analysis of the isolating oil. Although their generalized use, the diagnosis made by these techniques presents deficiencies, demanding the presence of specialists to complete the diagnosis. A great contribution for the electric sector would be a decision support tool capable of providing a correct and automatic diagnosis, to improve the monitoring process of power transformers. This article presents a diagnosis system, based on two artificial neural networks, each dedicated to the analysis of gaseous chromatography and physical-chemical of the isolating oil, respectively. The idea to enclose these two techniques is to accomplish a more complete diagnosis of the equipment, as well as a reduction of specialists' participation, creating a more automatic diagnosis system. The obtained results with the proposed system are compared with traditional methods. The resultant system represents a more complete decision support tool in the determination of the diagnosis of power transformers.

This paper presents a new methodology for meta-evaluation that makes use of fuzzy sets and fuzzy logic concepts. It comprehends a data collection instrument and a hierarchical fuzzy inference system. The advantages of the proposed system are: (i) the instrument, which allows intermediate answers; (ii) the inference process ability to adapt to specific needs; (ii) transparency, through the use of linguistic rules that helps both the understanding and the discussion of the whole process. The rules are based on guidelines established by the Joint Committee on Standards for Educational Evaluation and also represent the view of experts. The system can provide support to evaluators that may lack experience in meta-evaluation, which is the case in some developing countries. A case study is presented as a validation of the proposed methodology.

Since they were proposed as an optimization method, evolutionary algorithms(EA) have been used to solve problems in several research fields. This success is due, besides other things, to the fact that these algorithms do not require previous considerations regarding the problem to be optimized and offers a high degree of parallelism. However, some problems are computationally intensive regarding solution’s evaluation, which makes the optimization by EA’s slow for some situations. This paper proposes a novel EA for numerical optimiza tion inspired by the multiple universes principle of quantum computing. Results show that this algorithm can find better solutions, with less evaluations, when compared with similar algorithms.

This paper introduces the Inverted Hierarchical Neuro-Fuzzy BSP System (HNFB/sup -1/), a new neuro-fuzzy model that has been specifically created for record classification and rule extraction in databases. The HNFB/sup -1/ is based on the Hierarchical Neuro-Fuzzy Binary Space Partitioning Model (HNFB), which embodies a recursive partitioning of the input space, is able to automatically generate its own structure, and allows a greater number of inputs. The new HNFB/sup -1/ allows the extraction of knowledge in the form of interpretable fuzzy rules expressed by the following: If x is A and y is B, then input pattern belongs to class Z. For the process of rule extraction in the HNFB/sup -1/ model, two fuzzy evaluation measures were defined: 1) fuzzy accuracy and 2) fuzzy coverage. The HNFB/sup -1/ has been evaluated with different benchmark databases for the classification task: Iris Dataset, Wine Data, Pima Indians Diabetes Database, Bupa Liver Disorders, and Heart Disease. When compared with several other pattern classification models and algorithms, the HNFB/sup -1/ model has shown similar or better classification performance. Nevertheless, its performance in terms of processing time is remarkable. The HNFB/sup -1/ converged in less than one minute for all the databases described in the case study.

This paper proposes a fuzzy rule extraction method from support vector machines. Support vector machines (SVM) are learning systems based on statistical learning theory that have been successfully applied to a wide variety of application. However, SVM are "black box" models, that is, they generate a solution with linear combination of kernel functions which has a quite difficult interpretation. Methods for rule extraction from trained SVM have already been proposed, however, the rules generated by these methods have, in their antecedents, intervals or functions. This format decreases the interpretability of the generated rules and jeopardizes the knowledge extraction capability. Hence, to increase the linguistic interpretability of the generated rules, we propose in this paper a methodology for extracting fuzzy rules from a trained SVM, where the rule's antecedents are associated with fuzzy sets.

This work presents an extension of the hybrid reinforcement learning-hierarchical neuro-fuzzy politree model (RL-HNFP) and presents its performance in a multi-obstacle environment. The main objective of the RL-HNFP model is to provide an agent with intelligence, making it capable, by interacting with its environment, to acquire and retain knowledge for reasoning (infer an action). The original RL-HNFP applies hierarchical partitioning methods, together with the reinforcement learning (RL) methodology, which permits the autonomous agent to automatically learn its structure and its necessary action in each position in the environment. The improved version of the RL-HNFP model implements a better defuzzification method, improving the agent's behaviour. The extended RL-HNFP model was evaluated in a multi-obstacle environment, providing good performance and demonstrating the agent's autonomy.

This paper presents a neuro-fuzzy system developed to predict and classify the behaviour of steel beam subjected to concentrated loads. A good performance was obtained with a previously developed neural network system by Fonseca et al., (1999, 2001, 2003) when compared to available experimental data. The neural network accuracy was also significantly better than existing prediction formulae (Lyse and Godfrey, 1935; Bergfelt, 1979; Skaloud and Drdacky, 1975; Roberts and Newark, 1997). Despite this fact, the system architecture did not explicitly consider the different structural behaviour related to the beam collapse (web and flange yielding, web buckling and web crippling). Therefore this paper presents a neuro-fuzzy system that takes into account the ultimate limit state. The neuro-fuzzy system architecture is composed of one neuro-fuzzy model and one prediction neural network. The neuro-fuzzy model is used to classify the beams according to its pertinence to a specific structural response. Then, a neural network uses the pertinence established by the neuro-fuzzy classification model, to finally determine the beam patch load resistance.

This work presents the development of a methodology based on Monte Carlo simulation, fuzzy numbers and in the real options theory to determine the real options value under technical and market uncertainties. The objective of the proposed methodology is to substantially reduce the computational time involved, facilitating the decision taking process. The methodology involves: fuzzy numbers, to represent certain types of uncertainties that does not have a known stochastic process that can correctly model them; stochastic processes to represent other uncertainties; and Monte Carlo simulation to generate a good approximation of the real option value. This methodology was evaluated in problems of expansion option in the area of oil exploration and production, attaining the same results provided by conventional techniques but with a significant reduction in the necessary computational time.

This work presents a new hybrid neuro-fuzzy model for automatic learning of actions taken by agents. The main objective of this new model is to provide an agent with intelligence, making it capable, by interacting with its environment, to acquire and retain knowledge for reasoning (infer an action). This new model, named reinforcement learning hierarchical neuro-fuzzy politree (RL-HNFP), descends from the reinforcement learing hierarchical neuro-fuzzy BSP (RL-HNFB) that uses binary space partitioning. By using hierarchical partitioning methods, together with the reinforcement learning (RL) methodology, a new class of neuro-fuzzy systems (SNF) was obtained, which executes, in addition to automatically learning its structure, the autonomous learning of the actions to be taken by an agent. These characteristics represent an important differential when compared with the existing intelligent agents learning systems. The obtained results demonstrate the potential of this new model, which operates without any prior information, such as number of rules, rules specification, or number of partitions that the input space should have.

Additive models have been the preferential choice in nonlinear modeling: parametric or nonparametric, of conditional mean or variance. A new class of nonlinear additive varying coefficient models is presented in this paper. The coefficients are modeled by neural networks (multilayer perceptrons) and, both the conditional mean and conditional variance, are explicitly modeled. The learning algorithm of the neural network is based on a concept of likelihood maximization. Case studies with a nonlinear in variance synthetic series and a non-linear in mean real series are presented.

This work discusses an Evolvable Hardware (EHW) platform for the synthesis of analog electronic circuits. The EHW analog platform, named PAMA (Programmable Analog Multiplexer Array), is a reconfigurable platform that consists of integrated circuits whose internal connections can be programmed by Evolutionary Computation techniques, such as Genetic Algorithms, to synthesize circuits. The PAMA is classified as Field Programmable Analog Array (FPAA). FPAAs have just recently appeared, and most projects are being carried out in universities and research centers. They constitute the state of the art in the technology of reconfigurable platforms. These devices will become the building blocks of a forthcoming class of hardware, with the important features of self-adaptation and self-repairing, through automatic reconfiguration. The PAMA platform architectural details, concepts and characteristics are discussed. Three case studies, with promising results, are described: an operational amplifier, a logarithmic amplifier and a membership function circuit of a fuzzy logic controller.

This paper reviews the multiobjective fitness evaluation method called energy minimization (Zebulum et al., 1998; 2000), and presents an analysis of the method's behavior when used in a genetic algorithm applied to production scheduling of a petroleum refinery. The experimental results are presented and analyzed, leading to an overall evaluation of the benefits provided by the model.

This work investigates the application of variable length representation (VLR) evolutionary algorithms (EAs) in the field of Evolutionary Electronics. We propose a number of VLR methodologies that can cope with the main issues of variable length evolutionary systems. These issues include the search for efficient ways of sampling a genome space with varying dimensionalities, the task of balancing accuracy and parsimony of the solutions, and the manipulation of non-coding segments. We compare the performance of three proposed VLR approaches to sample the genome space: Increasing Length Genotypes, Oscillating Length Genotypes, and Uniformly Distributed Initial Population strategies. The advantages of reusing genetic material to replace non-coding segments are also emphasized in this work. It is shown, through examples in both analog and digital electronics, that the variable length genotype's representation is natural to this particular domain of application. A brief discussion on biological genome evolution is also provided.