Fidelis de Castro - Academia.edu (original) (raw)

Uploads

Papers by Fidelis de Castro

Research paper thumbnail of Identificação de Relevância em Textos de Sistemas de Help Desk usando Técnicas Clássicas de Aprendizado de Máquina

Procedings do XXII Congresso Brasileiro de Automatica, Oct 18, 2022

Research paper thumbnail of Relevance classification on service desk texts using Natural Language Processing

Service desk systems have a vast and rich base of information, consisting of the history of calls... more Service desk systems have a vast and rich base of information, consisting of the history of calls made, which can and should be used as a reference base for subsequent calls. Common search tools, such as keyword searches, prove to be unfeasible in large datasets, in addition to being able to bring results not necessarily related to the problem. “State-of-the-art” techniques exist, but they require high computational and operational costs for their training and use. In this sense, the purpose of this work is to investigate the sensitivity of machine learning algorithms in finding the characteristic defined here as “relevance”: the characteristic of texts with knowledge that can be reused. The motivation is that non-relevant texts can be removed in advance from the database, allowing complex algorithms to be employed in a more condensed database, reducing computational costs. Tests were performed with several combinations between the TF-IDF vectorizer and the word embedding Doc2Vec an...

Research paper thumbnail of Pré-diagnóstico de câncer de mama usando imagens histopatológicas com pré-processamento morfológico por meio de algoritmos clássicos e redes neurais profundas

Research paper thumbnail of Análise De Técnicas De Aprendizado De Máquina Aplicados Á Classificação De Grãos De Café

Ciências exatas e da terra: Conhecimentos didático-pedagógicos e o ensino-aprendizagem

Research paper thumbnail of Interações Automatizadas para Apoio à Modelagem de Processos de Negócio

Procedings do XV Simpósio Brasileiro de Automação Inteligente, 2021

Business Process Modeling is an activity increasingly adopted by organizations seeking to improve... more Business Process Modeling is an activity increasingly adopted by organizations seeking to improve their operational performance. A common practice for this activity is to performe interviews and workshops by involving modeling professionals (modelers) and domain experts. The produced models, in this case, are built based on the interpretation of the views gathered from domain experts. Considering this context, this paper discusses an approach for automated interactions between a computational agent and domain experts with the aim of determining process' control flow constraints, preventing potential inconsistencies and identifying modeling alternatives. The interaction generation mechanism, which is based on the building of the precess's discernment structure, provides to the computational agent the complete view of the modeling solution space. As a proof of concept, a prototype to support interactions was developed. The performed tests and experiments demonstrate that interactions produce valid and consistent models and allow the analysis of modeling alternatives. Resumo: A Modelagem de Processos de Negócioé uma atividade cada vez mais adotada por organizações que buscam melhorar seu desempenho operacional. Uma prática comum para esta atividade consiste na identificação do comportamento dos processos e sua transcrição em modelos, ao longo de entrevistas e oficinas envolvendo profissionais de modelagem (modeladores) e especialistas de domínio. Os modelos produzidos, neste caso, são construídos por modeladores tendo como base sua interpretação das visões colhidas junto a especialistas de domínio. Considerando este contexto, este trabalho discute um modelo de interações automatizadas com especialistas de domínio para capturar restrições de controle do fluxo do processo. O modelo concebido para apoiar as interações baseou-se na construção de uma estrutura de discernimento, que fornece a visão de todo o Espaço de Solução para o processo. Como prova de conceito foi desenvolvido um protótipo para interações entre um agente computacional e um humano. Conforme verificado em testes e experimentos com o protótipo, as interações permitem a análise de alternativas de modelos, produzem modelos consistentes e avaliam todo o Espaço de Solução de modelagem do processo.

Research paper thumbnail of Stability analysis of hypercomplex-valued discrete-time Hopfield-type neural networks

Research paper thumbnail of Continuous-Valued Octonionic Hopfield Neural Network

Proceeding Series of the Brazilian Society of Computational and Applied Mathematics, 2018

In this paper, we generalize the famous Hopfield neural network to unit octonions. In the propose... more In this paper, we generalize the famous Hopfield neural network to unit octonions. In the proposed model, referred to as the continuous-valued octonionic Hopfield neural network (CV-OHNN), the next state of a neuron is obtained by setting its octonionic activation potential to length one. We show that, like the traditional Hopfield network, a CV-OHNN operating in an asynchronous update mode always settles down to an equilibrium state under mild conditions on the octonionic synaptic weights.

Research paper thumbnail of Universidade Federal Do Espírito Santo Departamento De Matemática Programa De Mestrado Profissional Em Matemática Profmat

This work shows a proposal to approach Graph Theory, which is rarely taught at public high school... more This work shows a proposal to approach Graph Theory, which is rarely taught at public high schools, and looks for working the theme through the resolution of problems, providing opportunities to the pupil for effective participation on the building of arguments and challenging them to search for solutions, instigating the curiosity and requiring from them an attitude that lead to take decisions, favoring the emergence of creative answers and developing abilities concerning to those proposed on the Common Basic Curriculum of the state public high school. This proposal is described in two chapters, being the first one to introduce the theme on the second grade of high school, and the second one, to retake concepts already seen on the first one and introduce new concepts to the students of the third grade of high school. Both chapters are composed of problems, to introduce concepts and results as much to apply them, bringing their solutions and, in some cases, commentaries to the teach...

Research paper thumbnail of A broad class of discrete-time hypercomplex-valued Hopfield neural networks

Neural Networks, 2019

In this paper, we address the stability of a broad class of discrete-time hypercomplexvalued Hopf... more In this paper, we address the stability of a broad class of discrete-time hypercomplexvalued Hopfield-type neural networks. To ensure the neural networks belonging to this class always settle down at a stationary state, we introduce novel hypercomplex number systems referred to as real-part associative hypercomplex number systems. Real-part associative hypercomplex number systems generalize the well-known Cayley-Dickson algebras and real Clifford algebras and include the systems of real numbers, complex numbers, dual numbers, hyperbolic numbers, quaternions, tessarines, and octonions as particular instances. Apart from the novel hypercomplex number systems, we introduce a family of hypercomplex-valued activation functions called B-projection functions. Broadly speaking, a B-projection function projects the activation potential onto the set of all possible states of a hypercomplex-valued neuron. Using the theory presented in this paper, we confirm the stability analysis of several discrete-time hypercomplex-valued Hopfield-type neural networks from the literature. Moreover, we introduce and provide the stability analysis of a general class of Hopfield-type neural networks on Cayley-Dickson algebras.

Research paper thumbnail of Some Remarks on the Stability of Discrete-Time Complex-Valued Multistate Hopfield Neural Networks

Proceeding Series of the Brazilian Society of Computational and Applied Mathematics, 2018

In this paper, we review three discrete-time complex-valued Hopfield neural networks (CvMHNNs) pr... more In this paper, we review three discrete-time complex-valued Hopfield neural networks (CvMHNNs) proposed recently in the literature. Contrary to what has been stated, we provide examples in which the sequences produced by these CvMHNN fails to converge under the usual conditions on the synaptic weight matrix, that is, the synaptic weight matrix is hermitian with non-negative diagonal elements. Furthermore, we present one CvMHNN model that always settle down to a stationary state under the usual conditions on the synaptic weights.

Research paper thumbnail of On the Dynamics of Hopfield Neural Networks on Unit Quaternions

IEEE Transactions on Neural Networks and Learning Systems, 2017

In this paper, we first address the dynamics of the elegant multi-valued quaternionic Hopfield ne... more In this paper, we first address the dynamics of the elegant multi-valued quaternionic Hopfield neural network (MV-QHNN) proposed by Minemoto and collaborators. Contrary to what was expected, we show that the MV-QHNN, as well as one of its variation, does not always come to rest at an equilibrium state under the usual conditions. In fact, we provide simple examples in which the network yields a periodic sequence of quaternionic state vectors. Afterward, we turn our attention to the continuousvalued quaternionic Hopfield neural network (CV-QHNN), which can be derived from the MV-QHNN by means of a limit process. The CV-QHNN can be implemented more easily than the MV-QHNN model. Furthermore, the asynchronous CV-QHNN always settles down into an equilibrium state under the usual conditions. Theoretical issues are all illustrated by examples in this paper.

Research paper thumbnail of Theoretical and computational aspects of quaternionic multivalued Hopfield neural networks

2016 International Joint Conference on Neural Networks (IJCNN), 2016

Basic Concepts and Notation 2 MV-QHNN of Isokawa et al. 3 MV-QHNN of Minemoto et al.

Research paper thumbnail of Uma Proposta De Sequência Didática Para Treinamento Olímpico Em Matemática

A elaboração do presente trabalho envolve muito mais do que a titulação de mestre. Ela faz parte ... more A elaboração do presente trabalho envolve muito mais do que a titulação de mestre. Ela faz parte de um projeto de vida, e, se materializa como uma síntese de minha formação profissional e na paixão pelo ensino de Matemática. Agradeço a Deus, fonte de toda a vida, que me deu a oportunidade de finalizar meu mestrado com êxito. À minha esposa Mauricina, que me apoiou e me deu forças nos momentos de dificuldade, sempre demonstrando temperança e amor. À Minha filha Larissa, que com sua inocência e singeleza, me motiva a ser uma pessoa melhor a cada dia. À minha filha Lilian, que mesmo antes de vir à luz, já me impulsiona a buscar a felicidade sempre mais e mais.

Research paper thumbnail of Continuous-Valued Quaternionic Hopfield Neural Network for Image Retrieval: A Color Space Study

2017 Brazilian Conference on Intelligent Systems (BRACIS), 2017

Research paper thumbnail of Identificação de Relevância em Textos de Sistemas de Help Desk usando Técnicas Clássicas de Aprendizado de Máquina

Procedings do XXII Congresso Brasileiro de Automatica, Oct 18, 2022

Research paper thumbnail of Relevance classification on service desk texts using Natural Language Processing

Service desk systems have a vast and rich base of information, consisting of the history of calls... more Service desk systems have a vast and rich base of information, consisting of the history of calls made, which can and should be used as a reference base for subsequent calls. Common search tools, such as keyword searches, prove to be unfeasible in large datasets, in addition to being able to bring results not necessarily related to the problem. “State-of-the-art” techniques exist, but they require high computational and operational costs for their training and use. In this sense, the purpose of this work is to investigate the sensitivity of machine learning algorithms in finding the characteristic defined here as “relevance”: the characteristic of texts with knowledge that can be reused. The motivation is that non-relevant texts can be removed in advance from the database, allowing complex algorithms to be employed in a more condensed database, reducing computational costs. Tests were performed with several combinations between the TF-IDF vectorizer and the word embedding Doc2Vec an...

Research paper thumbnail of Pré-diagnóstico de câncer de mama usando imagens histopatológicas com pré-processamento morfológico por meio de algoritmos clássicos e redes neurais profundas

Research paper thumbnail of Análise De Técnicas De Aprendizado De Máquina Aplicados Á Classificação De Grãos De Café

Ciências exatas e da terra: Conhecimentos didático-pedagógicos e o ensino-aprendizagem

Research paper thumbnail of Interações Automatizadas para Apoio à Modelagem de Processos de Negócio

Procedings do XV Simpósio Brasileiro de Automação Inteligente, 2021

Business Process Modeling is an activity increasingly adopted by organizations seeking to improve... more Business Process Modeling is an activity increasingly adopted by organizations seeking to improve their operational performance. A common practice for this activity is to performe interviews and workshops by involving modeling professionals (modelers) and domain experts. The produced models, in this case, are built based on the interpretation of the views gathered from domain experts. Considering this context, this paper discusses an approach for automated interactions between a computational agent and domain experts with the aim of determining process' control flow constraints, preventing potential inconsistencies and identifying modeling alternatives. The interaction generation mechanism, which is based on the building of the precess's discernment structure, provides to the computational agent the complete view of the modeling solution space. As a proof of concept, a prototype to support interactions was developed. The performed tests and experiments demonstrate that interactions produce valid and consistent models and allow the analysis of modeling alternatives. Resumo: A Modelagem de Processos de Negócioé uma atividade cada vez mais adotada por organizações que buscam melhorar seu desempenho operacional. Uma prática comum para esta atividade consiste na identificação do comportamento dos processos e sua transcrição em modelos, ao longo de entrevistas e oficinas envolvendo profissionais de modelagem (modeladores) e especialistas de domínio. Os modelos produzidos, neste caso, são construídos por modeladores tendo como base sua interpretação das visões colhidas junto a especialistas de domínio. Considerando este contexto, este trabalho discute um modelo de interações automatizadas com especialistas de domínio para capturar restrições de controle do fluxo do processo. O modelo concebido para apoiar as interações baseou-se na construção de uma estrutura de discernimento, que fornece a visão de todo o Espaço de Solução para o processo. Como prova de conceito foi desenvolvido um protótipo para interações entre um agente computacional e um humano. Conforme verificado em testes e experimentos com o protótipo, as interações permitem a análise de alternativas de modelos, produzem modelos consistentes e avaliam todo o Espaço de Solução de modelagem do processo.

Research paper thumbnail of Stability analysis of hypercomplex-valued discrete-time Hopfield-type neural networks

Research paper thumbnail of Continuous-Valued Octonionic Hopfield Neural Network

Proceeding Series of the Brazilian Society of Computational and Applied Mathematics, 2018

In this paper, we generalize the famous Hopfield neural network to unit octonions. In the propose... more In this paper, we generalize the famous Hopfield neural network to unit octonions. In the proposed model, referred to as the continuous-valued octonionic Hopfield neural network (CV-OHNN), the next state of a neuron is obtained by setting its octonionic activation potential to length one. We show that, like the traditional Hopfield network, a CV-OHNN operating in an asynchronous update mode always settles down to an equilibrium state under mild conditions on the octonionic synaptic weights.

Research paper thumbnail of Universidade Federal Do Espírito Santo Departamento De Matemática Programa De Mestrado Profissional Em Matemática Profmat

This work shows a proposal to approach Graph Theory, which is rarely taught at public high school... more This work shows a proposal to approach Graph Theory, which is rarely taught at public high schools, and looks for working the theme through the resolution of problems, providing opportunities to the pupil for effective participation on the building of arguments and challenging them to search for solutions, instigating the curiosity and requiring from them an attitude that lead to take decisions, favoring the emergence of creative answers and developing abilities concerning to those proposed on the Common Basic Curriculum of the state public high school. This proposal is described in two chapters, being the first one to introduce the theme on the second grade of high school, and the second one, to retake concepts already seen on the first one and introduce new concepts to the students of the third grade of high school. Both chapters are composed of problems, to introduce concepts and results as much to apply them, bringing their solutions and, in some cases, commentaries to the teach...

Research paper thumbnail of A broad class of discrete-time hypercomplex-valued Hopfield neural networks

Neural Networks, 2019

In this paper, we address the stability of a broad class of discrete-time hypercomplexvalued Hopf... more In this paper, we address the stability of a broad class of discrete-time hypercomplexvalued Hopfield-type neural networks. To ensure the neural networks belonging to this class always settle down at a stationary state, we introduce novel hypercomplex number systems referred to as real-part associative hypercomplex number systems. Real-part associative hypercomplex number systems generalize the well-known Cayley-Dickson algebras and real Clifford algebras and include the systems of real numbers, complex numbers, dual numbers, hyperbolic numbers, quaternions, tessarines, and octonions as particular instances. Apart from the novel hypercomplex number systems, we introduce a family of hypercomplex-valued activation functions called B-projection functions. Broadly speaking, a B-projection function projects the activation potential onto the set of all possible states of a hypercomplex-valued neuron. Using the theory presented in this paper, we confirm the stability analysis of several discrete-time hypercomplex-valued Hopfield-type neural networks from the literature. Moreover, we introduce and provide the stability analysis of a general class of Hopfield-type neural networks on Cayley-Dickson algebras.

Research paper thumbnail of Some Remarks on the Stability of Discrete-Time Complex-Valued Multistate Hopfield Neural Networks

Proceeding Series of the Brazilian Society of Computational and Applied Mathematics, 2018

In this paper, we review three discrete-time complex-valued Hopfield neural networks (CvMHNNs) pr... more In this paper, we review three discrete-time complex-valued Hopfield neural networks (CvMHNNs) proposed recently in the literature. Contrary to what has been stated, we provide examples in which the sequences produced by these CvMHNN fails to converge under the usual conditions on the synaptic weight matrix, that is, the synaptic weight matrix is hermitian with non-negative diagonal elements. Furthermore, we present one CvMHNN model that always settle down to a stationary state under the usual conditions on the synaptic weights.

Research paper thumbnail of On the Dynamics of Hopfield Neural Networks on Unit Quaternions

IEEE Transactions on Neural Networks and Learning Systems, 2017

In this paper, we first address the dynamics of the elegant multi-valued quaternionic Hopfield ne... more In this paper, we first address the dynamics of the elegant multi-valued quaternionic Hopfield neural network (MV-QHNN) proposed by Minemoto and collaborators. Contrary to what was expected, we show that the MV-QHNN, as well as one of its variation, does not always come to rest at an equilibrium state under the usual conditions. In fact, we provide simple examples in which the network yields a periodic sequence of quaternionic state vectors. Afterward, we turn our attention to the continuousvalued quaternionic Hopfield neural network (CV-QHNN), which can be derived from the MV-QHNN by means of a limit process. The CV-QHNN can be implemented more easily than the MV-QHNN model. Furthermore, the asynchronous CV-QHNN always settles down into an equilibrium state under the usual conditions. Theoretical issues are all illustrated by examples in this paper.

Research paper thumbnail of Theoretical and computational aspects of quaternionic multivalued Hopfield neural networks

2016 International Joint Conference on Neural Networks (IJCNN), 2016

Basic Concepts and Notation 2 MV-QHNN of Isokawa et al. 3 MV-QHNN of Minemoto et al.

Research paper thumbnail of Uma Proposta De Sequência Didática Para Treinamento Olímpico Em Matemática

A elaboração do presente trabalho envolve muito mais do que a titulação de mestre. Ela faz parte ... more A elaboração do presente trabalho envolve muito mais do que a titulação de mestre. Ela faz parte de um projeto de vida, e, se materializa como uma síntese de minha formação profissional e na paixão pelo ensino de Matemática. Agradeço a Deus, fonte de toda a vida, que me deu a oportunidade de finalizar meu mestrado com êxito. À minha esposa Mauricina, que me apoiou e me deu forças nos momentos de dificuldade, sempre demonstrando temperança e amor. À Minha filha Larissa, que com sua inocência e singeleza, me motiva a ser uma pessoa melhor a cada dia. À minha filha Lilian, que mesmo antes de vir à luz, já me impulsiona a buscar a felicidade sempre mais e mais.

Research paper thumbnail of Continuous-Valued Quaternionic Hopfield Neural Network for Image Retrieval: A Color Space Study

2017 Brazilian Conference on Intelligent Systems (BRACIS), 2017