Guilherme A. Barreto

Universidade Federal do Ceará | UFC · Departamento de Engenharia de Teleinformática

Material didático (slides) usado nas disciplinas de Reconhecimento de Padrões e Inteligência Computacional sobre Análise de Componentes Principais.

Material didático (slides) usado nas disciplinas de Reconhecimento de Padrões e Inteligência Computacional sobre Clusterização de Dados usando métodos particionais (K-médias, K-médias com Covariância, K-medianas, K-medóides, Fuzzy K-médias).

Currently, the development of accurate and reliable models for predicting the behavior of rock mass joints is one of the most common interests among researchers, engineers and geologists. An alternative to address this type of problem more efficiently can be neuro-fuzzy systems, which combine the advantages of Fuzzy Controllers and Artificial Neura...

Artificial neural networks (ANNs) comprise parallel and distributed computational tools that can learn from data and make inferences (i.e., predictions) for highly nonlinear systems. By its turn, Petri nets (PNs) consist of well established modeling tools for parallel and distributed discrete event systems with a number of successful contributions...

The data were collected as the SCITOS G5 robot navigates through the room following the wall in a clockwise direction, for 4 rounds, using 24 ultrasound sensors arranged circularly around its 'waist'. The provided files comprise three different data sets. The first one contains the raw values of the measurements of all 24 ultrasound sensors and the...

Data set containing values for six biomechanical features used to classify orthopaedic patients into 3 classes (normal, disk hernia or spondilolysthesis) or 2 classes (normal or abnormal). Biomedical data set built by Dr. Henrique da Mota during a medical residence period in the Group of Applied Research in Orthopaedics (GARO) of the Centre Médico...

Water supply systems risk collapsing during droughts, which can affect millions of people. To mitigate these risks, we developed a proactive drought management system that integrates climate, hydrological variables, and mathematical modeling. The proposed Integrated Information and Early Warning System for Drought (IIEWSD) includes three main compo...

In this work, a fuzzy voltage controller design of a 1 kW high-gain, high-efficiency direct current converter operating in discontinuous conduction mode is developed. In this condition, the design of a conventional controller is more challenging. This converter is part of an autonomous photovoltaic pumping system without batteries consisting of fou...

This integrative review seeks to present an overview of the application of machine learning (ML) tools in the assessment of the risk of falls in the elderly. We searched the CAPES and IEEE Xplore Periodical databases, articles published in English, Portuguese and Spanish, in the last eleven years. Thirteen articles were selected. Most studies use d...

This integrative review seeks to present an overview of the application of machine learning (ML) tools in the assessment of the risk of falls in the elderly. We searched the CAPES and IEEE Xplore Periodical databases, articles published in English, Portuguese and Spanish, in the last eleven years. Thirteen articles were selected. Most studies use d...

Control systems of aircraft dynamics focus on choosing configurations that meet certain criteria related to speed response, damping ratio, error in steady state, among others. Aeronautical engineering applications have safety requirements, and in this respect, the use of optimization algorithms assist in the ability to develop models of increasing...

In this work, the identification of the thermal system of a neonatal incubator is performed using a local ARX LPV model whose scheduling variable is the ambient temperature. The LPV model performance indices are VAF=97.90% and RMSE=0.0077. The performance of this model is compared with ARX LTI models obtained by fixing the scheduling variable of th...

In this paper, we develop a self-growing variant of the local model network (LMN) for recursive dynamical system identification. The proposed model has the following features: growing online structure, fast recursive updating rules, better memory use (no storage of covariance matrices is required), and outlier-robustness. In this regard, efficiency...

In a supervised setting, the global classification paradigm leverages the whole training data to produce a single class discriminative model. Alternatively, the local classification approach builds multiple base classifiers, each of them using a small subset of the training data. In this paper, we take a path to stand in-between the global and loca...

A successful application of neural networks to the prediction of four important mechanical properties of steel rebar used in civil construction has been reported recently. In the current work, we advanced further in this issue by evaluating the performances of three kernel-based regression models, namely, the minimal learning machine (MLM), the sup...

Processing big data streams through machine learning algorithms has various challenges, such as little time to train the models, hardware memory constraints, and concept drift. In this paper, we show that prototype-based kernel classifiers designed by sparsification procedures, such as the approximate linear dependence (ALD) method, provides an ade...

Wind turbine power curve (WTPC) modeling from measured data is essential to predict the power generation from wind farms. Polynomial regression is commonly the first choice for this purpose, but there are other more powerful alternatives based on neural networks and fuzzy algorithms, for instance. Despite the existence of previous applications of s...

We introduce a new sparse variant of the Correntropy Kernel Learning model, hereafter named Fully ADaptive Online Sparse CKL (FADOS-CKL), for online system identification in the presence of outliers. For this purpose, we develop a fully adaptive dictionary of support vectors (SVs) so that it can either grow (as most of the kernel models to date do)...

In this paper, two contributions are presented. Firstly, a novel variant of the Kalman filter is developed and used to improve the tracking performance of the extended kernel recursive least squares algorithm (Ex-KRLS). Without resorting to the Riccati equation, the proposed formulation of the Kalman filter relies on principles of optimization and...

O presente trabalho aborda o processo de extração de atributos de sinais de eletroencefalograma (EEG), também chamado de parametrização de sinais, para o auxílio na detecção de crises epiléticas. A abordagem proposta neste trabalho é baseada na utilização de matrizes kernel para parametrizar segmentos do sinal de EEG de modo a construir vetores de...

A determinação do risco de cair é de suma importância na assistência à saúde do idoso, pois a ocorrência de quedas nessa população trazem consequências em vários aspectos. Ferramentas de aprendizado de máquinas têm sido cada vez mais empregadas com este fim. Portanto, o objetivo deste estudo foi investigar a viabilidade da utilização de sinais elet...

O presente trabalho apresenta um estudo sobre análise das atividades de sinais de Eletromiografia (EMG) a fim de realizar o reconhecimento de cinco gestos faciais. Em uma etapa inicial, é realizada a aquisição desses sinais estudando a escolha das expressões faciais e um melhor posicionamento dos sensores. Na etapa de classificação, utilizaram-se c...

Neste trabalho, é abordado um projeto de controlador fuzzy de tensão de um conversor CC-CC de alto ganho e alto rendimento de 1 kW operando em modo de condução descontínua. Este modo de condução para este conversor específico dificulta a utilização de controladores clássicos. O conversor em questão faz parte de um sistema de bombeamento fotovoltaic...

Este trabalho trata do reconhecimento de comandos de voz para o acionamento de um robô móvel. Comandos básicos formados pela elocução das palavras avançar, direita, esquerda, parar e recuar são gravados por um usuário e utilizados para criação de um banco de arquivos de áudio. A partir dos áudios gravados, técnicas de extração de atributos de sinai...

Data from real-world regression problems are quite often contaminated with outliers. In order to efficiently handle such undesirable samples, robust parameter estimation methods have been incorporated into randomized neural network (RNN) models, usually replacing the ordinary least squares (OLS) method. Despite recent successful applications to out...

The minimum jerk principle is commonly used for trajectory planning of robotic manipulators. However, since this principle is stated in terms of the robot’s kinematics, there is no guarantee that the joint controllers will actually track the planned acceleration and jerk profiles because the tuning of the controllers’ gains is decoupled from the tr...

The wind turbine power curve (WTPC) is a mathematical model built to capture the input-output relationship between the generated electrical power and the wind speed. An adequately fitted WTPC aids in wind energy assessment and prediction since the actual power curve will differ from that provided by the manufacturer due to a variety of reasons, suc...

The evaluation of finished (i.e., chemically treated) goat/sheep leather can be highly subjective, resulting in disagreements that can eventually lead to the interruption of production programs in the tannery and leather industry. As a result, much research has been carried out in the leather industry aiming at developing an automated system to eva...

The approximate linear dependence (ALD) method is a sparsification procedure used to build a dictionary of samples extracted from a dataset. The extracted samples are approximately linearly independent in a high-dimensional kernel reproducing Hilbert space. In this paper, we argue that the ALD method itself can be used to select relevant prototypes...

In this paper, we introduce a design methodology for prototype-based classifiers, more specifically the well-known LVQ family, aiming at improving their accuracy in fault detection/classification tasks. A laboratory testbed is constructed to generate the datasets which are comprised of short-circuit faults of different impedance levels, in addition...

A epilepsia é um distúrbio neurológico caracterizado por uma perturbação elétrica anormal no cérebro, causando convulsões recorrentes. O exame mais utilizado no diagnóstico da epilepsia é o eletroencefalograma (EEG), onde a atividade elétrica cerebral de um paciente é mensurada e analisada visualmente. Contudo, identificar os padrões epilépticos no...

The aim of this paper is to present a practical application of Fuzzy Controllers and Artificial Neural Networks (ANN) models to analyze the stability of rock slopes defined by unfilled discontinuities. Results from direct shear tests on different boundary conditions and types of discontinuities have been used to develop these models based on the co...

The present study evaluates the viscera of the Nile tilapia (bred in captivity) for bio-oil production and its impact on the biodiesel supply chain. We report in detail all the steps, from viscera extraction to the economic viability of the produced biodiesel, that led to the development of an oil extraction unit and the separation of the resulting...

The use of recurrent neural networks in online system identification is very limited in real-world applications, mainly due to the propagation of errors caused by the iterative nature of the prediction task over multiple steps ahead. Bearing this in mind, in this paper, we revisit design issues regarding the robustness of the echo state network (ES...

It is well known that learning about the mechanical behavior of a fractured rock mass depends on the shear behavior of its discontinuities. Several studies have shown that the shear behavior of unfilled rock discontinuities depends on their boundary conditions, roughness characteristics and the properties of the joints walls. Currently, there are s...

As a result of a number of studies, some analytical models have been developed to predict the shear behavior of unfilled rock joints, but they all present a purely deterministic nature because their input variables are defined without considering the uncertainties inherent in the formation processes of the rock masses and related discontinuities. T...

Gaussian Processes (GPs) models have been successfully applied to the problem of learning from sequential observations. In such context, the family of Recurrent Gaussian Processes (RGPs) have been recently introduced with a specifically designed structure to handle dynamical data. However, RGPs present a limitation shared by most GP approaches: the...

In this paper, we introduce a sparse variant of the Correntropy Kernel Learning (CKL) model for online system identification in the presence of outliers. The proposed Online Sparse CKL(OS-CKL) improves the original CKL in three important aspects. Firstly, it is modified to operate as a kernel adaptive filter, i.e. model-building is a continuous pro...

In this paper we revisit the design of neural-network based local linear models for dynamic system identification aiming at extending their use to scenarios contaminated with outliers. To this purpose, we modify well-known local linear models by replacing their original recursive rules with outlier-robust variants developed from the M-estimation fr...

In this paper we introduce a novel technique for optimal tuning of PD controllers engaged in tracking minimum-jerk (MJ) trajectories. The proposed approach is an attempt to bridge the gap between the MJ principle for trajectory planning, which is based solely on the robot's kinematics, and the optimal estimation of the gains of the joint controller...

In this paper, the classical polynomial model for wind turbines power curve estimation is revisited aiming at an automatic and parsimonious design. In this regard, using genetic algorithms we introduce a methodoloy for estimating a suitable order for the polynomial as well its relevant terms. The proposed methodology is compared with the state of t...

Sabe-se que o conhecimento do comportamento mecânico de um maciço rochoso com certo grau de fraturamento está, diretamente, vinculado às características de suas descontinuidades e à resposta das mesmas às solicitações atuantes. Diante disso, torna-se fundamental a utilização dos parâmetros adequados para a correta previsão do comportamento das desc...

The permanent magnet (PM) spherical motor has been subject to growing interest from the scientific community due to its potential for applications in distinct areas, particularly in robotics, prosthetics, satellite control, sensors or camera systems. Motivated by this movement, the current work presents all the steps for the efficient design and co...

O conhecimento do comportamento mecânico de um maciço rochoso com certo grau de fraturamento está, diretamente, vinculado com as características de suas descontinuidades e a resposta da mesma às solicitações atuantes. Diante disso, para a análise de taludes rochosos por exemplo, torna-se fundamental a utilização dos parâmetros adequados para a corr...

In this survey paper, we report the results of a comprehensive study involving the application of dynamic self-organizing neural networks (SONNs) to the problem of novelty detection in time series data. The study is comprised of three main parts. In the first part, we aim at evaluating how the performances of nonrecurrent dynamic SONNs are influenc...

In this paper we revisit the design of local linear models for dynamic system identification aiming at adapting their use to scenarios contaminated with outliers. To this purpose, we evaluate the performance of three local linear models originated within the field of artificial neural networks and introduce modifications in their learning rules wit...

This book constitutes the thoroughly refereed proceedings of the 37th IFSA Conference, NAFIPS 2018, held in Fortaleza, Brazil, in July 2018. The 55 full papers presented were carefully reviewed and selected from 73 submissions. The papers deal with a large spectrum of topics, including theory and applications of fuzzy numbers and sets, fuzzy logi...

The kernel recursive least squares (KRLS), a nonlinear counterpart of the famed RLS algorithm, performs linear regression in a high-dimensional feature space induced by a Mercer kernel. Despite the growing interest in the KRLS for nonlinear signal processing, the presence of outliers in the estimation data causes the resulting predictor’s performan...

In this work the characterization of the Nile Tilapia viscera was performed. To do this, the oil (lipids) and the dreg (protein) were extracted. Acidity parameters were evaluated in the presence and absence of biliary juice. The other byproduct analyzed, dreg (protein), was evaluated in terms of its potential use as a raw material in the production...

This paper relies on the resource-oriented architecture (ROA) to propose a middleware that shares resources (sensors, actuators and services) of one or more robots through the TCP/IP network, providing greater efficiency in the development of software applications for robotics. The proposed middleware consists of a set of web services that provides...

Gaussian Processes (GP) comprise a powerful kernel-based machine learning paradigm which has recently attracted the attention of the nonlinear system identification community, specially due to its inherent Bayesian-style treatment of the uncertainty. However, since standard GP models assume a Gaussian distribution for the observation noise, i.e., a...

The Forward Stagewise Regression (FSR) algorithm is a popular procedure to generate sparse linear regression models. However, the standard FSR assumes that the data are fully observed. This assumption is often flawed and pre-processing steps are applied to the dataset so that FSR can be used. In this paper, we extend the FSR algorithm to directly h...

In this paper, randomized single-hidden layer feedforward networks (SLFNs) are extended to handle outliers sequentially in online system identification tasks involving large-scale datasets. Starting from the description of the original batch learning algorithms of the evaluated randomized SLFNs, we discuss how these neural architectures can be easi...

Recent demands from big data applications have strongly motivated a successful sparse formulation of the least squares support vector regression (LSSVR) model in primal weight space. Such an approach, which has been called fixed-size LSSVR (FS-LSSVR), is built upon an approximation of the nonlinear feature mapping based on Nyström method to overcom...