Bauru - FC - Faculdade de Ciências

URI Permanente para esta coleçãohttps://hdl.handle.net/11449/253805

Navegar

Agora exibindo 1 - 20 de 29

Harmony Search-Based Approaches for Fine-Tuning Deep Belief Networks
(2023-01-01) Rodrigues, Douglas ; Roder, Mateus ; Passos, Leandro Aparecido ; Rosa, Gustavo Henrique de ; Papa, João Paulo ; Geem, Zong Woo ; Universidade Estadual Paulista (UNESP) ; University of Wolverhampton ; Gachon University
Harmony Search (HS) is a metaheuristic algorithm inspired by the musical composition process, precisely the composition of harmonies, i.e., the chain of different musical notes. The algorithm’s simplicity allows several points to improve to explore the entire search space efficiently. This work aims to compare different HS variants in image restoration using Deep Belief Networks (DBN). We compared standard HS against five variants: Improved Harmony Search (IHS), Self-adaptive Global Best Harmony Search (SGHS), Global-best Harmony Search (GHS), Novel Global Harmony Search (NGHS), and Global Harmony Search with Generalized Opposition-based learning (GOGHS). Experiments in public datasets for binary image reconstruction highlighted that HS and its variants obtained superior results than a random search used as a baseline. Also, it was found that the GHS variant is inferior to the others for some cases.
Basic concepts of behavioral pharmacology
(2021-09-23) Gonçãlves, Fábio Leyser ; Universidade Estadual Paulista (UNESP) ; Universidade de São Paulo (USP)
This chapter intends to present some concepts, methods, and contributions of Behavioral Pharmacology to the understanding of the phenomenon of substance dependence. The chapter begins with a brief description of the area and history of behavioral pharmacology. Next, it presents basic concepts of pharmacology and neurobiological processes involved in the action of psychotropic substances, seeking to help the reader to develop a basic repertoire that helps him/her to face the challenges of an interdisciplinary area. Among the main concepts approached are absorption, distribution, and excretion of substances by the body. On the sequence, it seeks to clarify the basic principles of the neurotransmission chemical process, the basis for the functioning of the entire nervous system. Some of the main methods of behavioral pharmacology are explored, with emphasis on those related to substance dependence, such as self-administration procedures, second-order schedules of reinforcement and procedures related to conditioned reinforcement, concurrent schedules choice and behavioral economics. The chapter also discusses a proposal for understanding the relationship between drugs and behavior, including organism modifications related to the effects of psychotropic substances. Finally, some information is briefly presented on three drugs commonly related to the dependence phenomenon, especially in Brazil: alcohol, nicotine, and cocaine.
Future trends in optimum-path forest classification
(2022-01-24) Papa, João Paulo ; Falcão, Alexandre Xavier ; Universidade Estadual Paulista (UNESP) ; Universidade Estadual de Campinas (UNICAMP)
In the past years, we have observed an increasing number of applications that require machine learning techniques to sort out problems that are not straightforward to humans. The reasons vary from information that is not clearly visible to the human eye (e.g., microscopic patterns in medical images) or the massive amount of data to analyze. This book aimed to shed light on the Optimum-Path Forest framework, which comprises approaches to dealing with supervised, semi-supervised, and unsupervised learning. Different applications have been presented together with a theoretical background concerning the techniques presented here. We expect to call the attention and curiosity of the readers towards OPF-based techniques and their strengths. © 2022 Copyright
Theoretical background and related works
(2022-01-24) Afonso, Luis C.S. ; Falcão, Alexandre Xavier ; Papa, João Paulo ; Universidade Estadual Paulista (UNESP) ; Universidade Estadual de Campinas (UNICAMP)
The Optimum-Path Forest (OPF) is a framework for the design of graph-based classifiers, which covers supervised, semisupervised, and unsupervised applications. The OPF is mainly characterized by its low training and classification times as well as competitive results against well-established machine learning techniques, such as Support Vector Machine and Artificial Neural Networks. Besides, the framework allows the design of different approaches based on the problem itself, which means a specific OPF-based classifier can be built for a given particular task. This paper surveyed several works published in the past years concerning OPF-based classifiers and sheds light on future trends concerning such a framework in the context of the deep learning era. © 2022 Copyright
Learning to weight similarity measures with Siamese networks: A case study on optimum-path forest
(2022-01-24) De Rosa, Gustavo H. ; Papa, João Paulo ; Universidade Estadual Paulista (UNESP)
Recent advances in machine learning algorithms have been aiding humans and improving their decision-making capacities in various applications, such as medical imaging, image classification and reconstruction, object recognition, and text categorization. A graph-based classifier, known as Optimum-Path Forest (OPF), has been extensively researched in the last years, mainly due to its parameterless nature and state-of-the-art results compared to well-known literature classifiers, for example, support vector machines. Nevertheless, one drawback concerning such an approach lies in its distance calculation, which has to be selected from a range of formulae and computed between all nodes to weigh the graph's arcs, and hence time-consuming. Therefore in this work, we propose to address such a problem by precomputing the arcs' distances through a similarity measure obtained from Siamese networks. The idea is to employ the same training set used by the OPF classifier to train a Siamese network and calculate the samples' distance through a similarity measure. The experimental results show that the proposed method is suitable, where the similarity-based OPF achieved comparable results to its standard counterpart and even surpassed it in some datasets. Additionally, the precalculated similarity matrix lessens the burden of recalculating the distances for every new classification. © 2022 Copyright
Detecting atherosclerotic plaque calcifications of the carotid artery through optimum-path forest
(2022-01-24) Jodas, Danilo Samuel ; Roder, Mateus ; Pires, Rafael ; Silva Santana, Marcos Cleison ; de Souza, Luis A. ; Passos, Leandro Aparecido ; Universidade Estadual Paulista (UNESP) ; São Carlos Federal University
Analysis of the atherosclerotic lesions deposited in the carotid artery is so far an essential task for estimating possible cardiovascular disorders in patients. The immediate assessment of such lesion, as well as its morphology and composition, turns out necessary to avoid its progression beforehand, thus preventing more severe conditions such as heart attacks and strokes caused by calcified elements observed in advanced stages. Heretofore, a number of works addressed medical diagnosis problems through computational approaches, developing Computer-Aided Diagnosis (CAD) tools to detect, among several applications, atherosclerotic plaques formed in carotid arteries. In this context, a graph-based machine learning framework called Optimum-Path Forest (OPF) was successfully employed to tackle several CAD-based problems, even though no one still explores the model to classify the task mentioned above. Therefore this paper proposes the classification of regions in atherosclerotic lesions as calcified or noncalcified debris through OPF-based approaches. In the process, handcrafted features are extracted from pixels of computed tomography angiography images of the carotid artery. Also, each pixel is labeled by an expert as a calcified or noncalcified element. Thereafter, the OPF classifier, as well as four variants, namely Fuzzy OPF, OPF. knn, Probabilistic OPF, and the OPF for anomaly detection, are compared for the task of predicting whether the pixel of the carotid artery stands for the calcium of the atherosclerotic lesion or not. © 2022 Copyright
Convolutional neural networks applied for Parkinson’s disease identification
(2016-01-01) Pereira, Clayton R. ; Pereira, Danillo R. ; Papa, Joao P. ; Rosa, Gustavo H. ; Yang, Xin-She ; Universidade Federal de São Carlos (UFSCar) ; Universidade Estadual Paulista (UNESP) ; Middlesex University
Parkinson’s Disease (PD) is a chronic and progressive illness that affects hundreds of thousands of people worldwide. Although it is quite easy to identify someone affected by PD when the illness shows itself (e.g. tremors, slowness of movement and freezing-of-gait), most works have focused on studying the working mechanism of the disease in its very early stages. In such cases, drugs can be administered in order to increase the quality of life of the patients. Since the beginning, it is well-known that PD patients feature the micrography, which is related to muscle rigidity and tremors. As such, most exams to detect Parkinson’s Disease make use of handwritten assessment tools, where the individual is asked to perform some predefined tasks, such as drawing spirals and meanders on a template paper. Later, an expert analyses the drawings in order to classify the progressive of the disease. In this work, we are interested into aiding physicians in such task by means of machine learning techniques, which can learn proper information from digitized versions of the exams, and them recommending a probability of a given individual being affected by PD depending on its handwritten skills. Particularly, we are interested in deep learning techniques (i.e. Convolutional Neural Networks) due to their ability into learning features without human interaction. Additionally, we propose to fine-tune hyper-arameters of such techniques by means of meta-heuristic-based techniques, such as Bat Algorithm, Firefly Algorithm and Particle Swarm Optimization.
Normalizing images is good to improve computer-assisted COVID-19 diagnosis
(2021-01-01) Santos, Claudio Filipi Gonçalvesdos ; Passos, Leandro Aparecido ; Santana, Marcos Cleisonde ; Papa, João Paulo ; Universidade Federal de São Carlos (UFSCar) ; Universidade Estadual Paulista (UNESP)
The Coronavirus Disease 2019 (COVID-19) outbreak, caused by the SARS-CoV-2 virus, surprised the whole world in an unprecedented and devastating way, resulting in almost deaths and 2.3 million infections worldwide in less than 4 months. Moreover, the elevate capability of transmission threatens to collapse both the healthy and economic systems from most countries, stressing worse predictions for emerging countries. In such a turbulent scenario, fast diagnosis is essential for a successful treatment and isolation of patients, thus avoiding increasing the number of contaminations. However, traditional methods of detection using polymerase chain reaction are impractical in large scale due to elevate costs, material scarcity, and time demanded for processing. As an alternative, some researchers proposed a machine learning-based diagnosis considering chest X-ray analysis with promising results, thus opening room for possible improvements. This work introduces a different normalization approach that, together with an EfficientNet-B6-inspired neural network, can deal with COVID-19 diagnosis considering chest X-ray images. Experiments provided competitive results considering a lighter and faster architecture, thus fostering research toward COVID-19 detection.
Analysis of Pedagogical Practices Carried Out in Continuing Education Activities for Physics Teachers: Limits and Possibilities
(2018-01-01) Gatti, Sandra Regina Teodoro ; Nardi, Roberto ; Universidade Estadual Paulista (UNESP)
The chapter deals with a general methodological and epistemological contribution to continuing education activities for physics teachers. Researchers on science teaching seem to point to a consensus on the importance of the historical and philosophical approach to teaching. Nevertheless, the pedagogical practice developed in classroom hardly ever embodies such recommendations. In this research, the authors planned a continuing education course and followed the development of five in-service physics teachers during the school year of 2008. The researchers discussed with teachers, focusing on their experiences, and seeking to assist them in the construction of alternatives means of observing and understanding the students’ work. The authors seek to revisit the training actions by discussing the propositions and perceptions of two of the participants and the respective impacts of the experience, 3 years later, when one of them became a supervisor to pre-service teacher.
Recent advances on optimum-path forest for data classification: Supervised, semi-supervised, and unsupervised learning
(2015-12-15) Papa, João Paulo ; Amorim, Willian Paraguassu ; Falcão, Alexandre Xavier ; Tavares, João Manuel R.S. ; Universidade Estadual Paulista (UNESP) ; Federal University of Dourados Region ; Universidade Estadual de Campinas (UNICAMP) ; Universidade do Porto
Although one can find several pattern recognition techniques out there, there is still room for improvements and new approaches. In this book chapter, we revisited the Optimum-Path Forest (OPF) classifier, which has been evaluated over the last years in a number of applications that consider supervised, semi-supervised and unsupervised learning problems. We also presented a brief compilation of a number of previous works that employed OPF in different research fields, that range from remote sensing image classification to medical data analysis.
Solvent Effects on Electronic Circular Dichroism Spectra
(2017-01-01) De Souza, Aguinaldo Robinson ; Ximenes, Valdecir Farias ; Morgon, Nelson Henrique ; Universidade Estadual Paulista (UNESP) ; Universidade Estadual de Campinas (UNICAMP)
The Electronic Circular Dichroism Spectrum (ECD) is a valuable tool to study the unknown absolute configuration of an optically active molecule. And the comparison between experimental data and theoretical computational calculations has been a successful strategy for this study. However, the ECD spectrum is very sensitive to solvent effects that significantly change the character of the results obtained. This chapter is focused on the study of the solvent effects and their application in both experimental and computational chemistry of ECD of the compound 3,3′-dibromo-1,1′-bi-2-naphthol.
On the Assessment of Nature-Inspired Meta-Heuristic Optimization Techniques to Fine-Tune Deep Belief Networks
(2020-01-01) Passos, Leandro Aparecido ; Rosa, Gustavo Henrique de ; Rodrigues, Douglas ; Roder, Mateus ; Papa, João Paulo ; Universidade Estadual Paulista (UNESP) ; São Carlos Federal University
Machine learning techniques are capable of talking, interpreting, creating, and even reasoning about virtually any subject. Also, their learning power has grown exponentially throughout the last years due to advances in hardware architecture. Nevertheless, most of these models still struggle regarding their practical usage since they require a proper selection of hyper-parameters, which are often empirically chosen. Such requirements are strengthened when concerning deep learning models, which commonly require a higher number of hyper-parameters. A collection of nature-inspired optimization techniques, known as meta-heuristics, arise as straightforward solutions to tackle such problems since they do not employ derivatives, thus alleviating their computational burden. Therefore, this work proposes a comparison among several meta-heuristic optimization techniques in the context of Deep Belief Networks hyper-parameter fine-tuning. An experimental setup was conducted over three public datasets in the task of binary image reconstruction and demonstrated consistent results, posing meta-heuristic techniques as a suitable alternative to the problem.
Adaptive improved flower pollination algorithm for global optimization
(2020-01-01) Rodrigues, Douglas ; de Rosa, Gustavo Henrique ; Passos, Leandro Aparecido ; Papa, João Paulo ; São Carlos Federal University ; Universidade Estadual Paulista (UNESP)
In the last few years, meta-heuristic-driven optimization algorithms have been employed to solve several problems since they can provide simple and elegant solutions. In this work, we introduced an improved adaptive version of the Flower Pollination Algorithm, which can dynamically change its parameter setting throughout the convergence process, as well as it keeps track of the best solutions. The effectiveness of the proposed approach is compared against with Bat Algorithm and Particle Swarm Optimization, as well as the naïve version of the Flower Pollination Algorithm. The experimental results were carried out in nine benchmark functions available in literature and demonstrated to outperform the other techniques with faster convergence rate.
On the hypercomplex-based search spaces for optimization purposes
(2018-01-01) Papa, João Paulo ; de Rosa, Gustavo Henrique ; Yang, Xin-She ; Universidade Estadual Paulista (UNESP) ; Middlesex University London
Most applications can be modeled using real-valued algebra. Nevertheless, certain problems may be better addressed using different mathematical tools. In this context, complex numbers can be viewed as an alternative to standard algebra, where imaginary numbers allow a broader collection of tools to deal with different types of problems. In addition, hypercomplex numbers extend naïve complex algebra by means of additional imaginary numbers, such as quaternions and octonions. In this work, we will review the literature concerning hypercomplex spaces with an emphasis on the main concepts and fundamentals that build the quaternion and octonion algebra, and why they are interesting approaches that can overcome some potential drawbacks of certain optimization techniques. We show that quaternion- and octonion-based algebra can be used to different optimization problems, allowing smoother fitness landscapes and providing better results than those represented in standard search spaces.
Marxist methodological foundations in Vygotsky’s work
(2017-01-01) Martins, Lígia Márcia ; Universidade Estadual Paulista (UNESP)
Gait velocity, attention and exercise in Parkinson's disease
(2012-07-01) Gobbi, Lilian Teresa Bucken ; Vitorio, Rodrigo ; Teixeira-Arroyo, Claudia ; Lirani-Silva, Ellen ; Rinaldi, Natalia Madalena ; Barbieri, Fabio Augusto ; Pereira, Marcelo Pinto ; Santos, Paulo Cezar Rocha ; Batistela, Rosangela Alice ; Universidade Estadual Paulista (UNESP)
In addition to pharmacological therapy, different modes of interventions has been purposed and applied for people with Parkinson's disease (PD). The results of these interventions have been shown improvements in their quality of life and functional mobility, among others. This study aimed to identify the effects of a 6-month, multimodal exercise program on stride velocity and attentional outcomes in PD patients and to observe the relationship between these dependent variables. Twenty-one patients with idiopathic PD were assessed before and after the multimodal exercise program (three times a week, one hour per session). Outcome measures included the stride velocity of self-paced walking and score (number of right/wrong answers) in the attention test of Wechsler Adult Intelligence Scale (WAIS-III). Gait velocity and the number of errors at WAIS-III improved after the enrollment in the multimodal exercise program. It was observed a significant positive relationship between gait velocity and number of right answers at WAIS-III before the multimodal exercise program. These findings suggest that PD patients with higher attentional levels show less impairment in gait velocity. Although physical exercise was effective to improve motor and cognitive characteristics of PD patients, further study in the form of randomized controlled trial would be required to establish effectiveness of the multimodal exercise program. © 2012 Nova Science Publishers, Inc. All rights reserved.
A structure for knowledge management systems assessment and audit
(2010-01-01) Albino, Joao Pedro ; Reinhard, Nicolau ; Santana, Silvina ; Universidade Estadual Paulista (UNESP) ; Universidade de São Paulo (USP) ; University of Aveiro
Knowledge Management Systems seek to offer a framework to stimulate the sharing of the intellectual capital ofan organization so that the resources invested in time and technology can be effectively utilized. Recent research has shown that some businesses invest thousands of dollars to establish knowledge management processes in their organizations. Others are still in the initial phase of introduction, and many of them would like to embark on such projects. It can be observed, however, that the great majority of such initiatives have not delivered the returns hoped for, since the greatest emphasis is given to questions of technology and to the methodologies of KM projects. In this study, we callattention to an emerging problem which recent studies of the phenomenon of knowledge sharing have not sufficiently addressed: the difficulties and efforts of organizations in identifying their centers of knowledge, in developing and implementing KM projects, and in utilizing them effectively. Thus, the objective of this chapter is to propose a framework to evaluate the present state of an organization’s processes and activities and identify which information and communication technologies (ICT) are supporting these initiatives, with the intention of diagnosing its real need for KM. Another objective of this instrument is to create a base of knowledge, with all the evaluations undertaken in organizations in different sectors and areas of specialization available to all participants in the process, as a way of sharing knowledge for continual improvement and dissemination of the best practices. About 30 companies took part in the first phase of investigation in 2008, and the knowledge base is under construction.
Bio-tribocorrosion in dental applications
(2013-01-01) Rocha, L. A. ; Oliveira, F. ; Cruz, H. V. ; Sukotjo, C. ; Mathew, M. T. ; Universidade Estadual Paulista (UNESP) ; Centre for Mechanical and Materials Technologies ; College of Dentistry (MC 555) ; Rush University Medical Center
In the oral cavity, materials (including our natural teeth) are exposed to a complex environment, which results in simultaneous mechanical, electrochemical, and microbiological solicitations. Therefore, bio-tribocorrosion is an important cause of degradation of dental materials leading to functional and/or biological detrimental effects due to an increased release of metallic ions and wear debris. This chapter describes the main bio-tribocorrosion phenomena that occur in the oral environment, and discuss the main parameters related to both the materials and the environment affecting bio-tribocorrosion in dental applications. © 2013 Woodhead Publishing Limited All rights reserved.
An overview on the photoluminescence emission in ZNO single crystal. A joint experimental and theoretical analysis
(2012-12-01) Lima, Renata C. ; Andrés, J. ; Sambrano, Júlio R. ; Longo, Valéria M. ; Macário, Leilane R. ; Espinosa, José W. M. ; Pizani, Paulo S. ; Varela, José A. ; Longo, Elson ; Universidade Federal de Uberlândia (UFU) ; Universitat Jaume I ; Universidade Estadual Paulista (UNESP) ; Universidade Federal de Goiás (UFG) ; Universidade Federal de São Carlos (UFSCar)
The purpose of the present chapter is to join both the experimental and theoretical results to explain the different responses of photoluminescence (PL) emission at room temperature of ZnO powders by using different structural order-disorder effects, associated with intermediate and short range defects. Recent theoretical and experimental analyzing the crystal and surface structures, electronic and PL properties ZnO-based materials have been presented. In addition, the synthetic methods, with particular attention to hydrothermal conventional and microwave assisted hydrothermal procedures, characterization techniques, as well as crystal growth process has been reviewed. The two synthesis methods used in the ZnO powders preparation, conventional hydrothermal and microwave hydrothermal, suggest different defects formation, which can be an indicative of different conformations into the hexagonal structure (wurtzite) during the ZnO growth. This review focuses on studies that employ electronic structure calculations, which primarily concentrate on using density functional theory (DFT) to understand PL behavior. DFT calculations, at the B3LYP level, have been combined with the results obtained by using X-ray diffraction, Raman, and photoluminescence techniques The theoretical results indicate that the key factor controlling the electronic behavior can be associated with a symmetry breaking process, creating localized electronic levels between the valence and conduction bands. These defects are associated structural disorder generated by the presence of distortions in the ideal constituent clusters of these materials [ZnO4] and [ZnO3. VzO]. © 2012 Nova Science Publishers, Inc.
A Hybrid Approach for Breast Mass Categorization
(2019-01-01) Passos, Leandro Aparecido ; Santos, Claudio ; Pereira, Clayton Reginaldo ; Afonso, Luis Claudio Sugi ; Papa, João P. ; Universidade Estadual Paulista (Unesp) ; Universidade Federal de São Carlos (UFSCar)
Breast cancer is one of the most frequent fatal diseases among women around the world. Early diagnosis is paramount for easing such statistics, increasing the probability of successful treatment and cure. This paper proposes a hybrid approach composed of a convolutional neural network with a supervised classifier on the top capable of predicting eight specific cases of the breast tumor, being four of them malignant and four benign. The model employs the BreastNet convolution neural network to the task of mammogram images feature extraction, and it compares three distinct supervised-learning algorithms for classification purposes: (i) Optimum-Path Forest, (ii) Support Vector Machines (SVM) with Radial Basis Function, and (iii) SVM with a linear kernel. Moreover, since BreastNet is also capable of performing classification tasks, its results are further compared against the other three techniques. Experimental results demonstrate the robustness of the model, achieving 86 % of accuracy over the public LAPIMO dataset.

Navegar

Submissões Recentes