Artigos - Estatistica, Matemática Aplicada e Computação - IGCE

URI Permanente para esta coleção

Navegar

Submissões Recentes

Agora exibindo 1 - 20 de 650
  • ItemArtigo
    CombTEs: combining predictions from the search for transposable elements
    (2022-01-01) Fischer, Carlos Norberto [UNESP]; Universidade Estadual Paulista (UNESP)
    Several tools, using different approaches, are available nowadays to identify transposable elements (TEs) in a query sequence. Normally, a same set of TEs can be predicted by many of these tools. However, for other TEs, only a few tools are able to predict them due to their particular characteristics. In both cases, combining predictions produced by two or more tools can be an interesting approach to increasing the number of correct results and, at the same time, to further improve the confidence about the predicted TEs. Taking this into account, this work presents an auxiliary tool, CombTEs, that combines predictions produced by other programs and pipelines used to identify TEs in a genome sequence. The basic idea is that, after running only once the tools of interest, the same sets of initial predictions are used in several combining processes, each one considering different values for the parameters used by CombTEs (for example, filters and distance between predictions), in a very fast way, making the annotation step easier and more reliable.
  • ItemArtigo
    Dynamic analysis of an electromagnetic vibration absorber in a non-ideal system
    (2023-01-01) Petrocino, Eduardo Abuhamad [UNESP]; Balthazar, José Manoel [UNESP]; Tusset, Ângelo Marcelo; Gonçalves, Paulo José Paupitz [UNESP]; Silveira, Marcos [UNESP]; Universidade Estadual Paulista (UNESP); Federal University of Technology – Paraná
    A dynamic system usually used in engineering are rotating machines fixed to flexible structures, and an excessive vibration known as the Sommerfeld effect may occur in the passage of the resonance. This research aims to analyze an electromagnetic absorber in a non-ideal system, a simple model for the analysis of the first mode of a beam. In this context, an investigation of the dynamic behavior of a cantilever beam with a dc motor coupled to its free end, having an unbalanced mass on its axis, providing a non-ideal energy source. A permanent magnet is attached to this same free end, which is immersed inside a coil attached to the base. In this step, an analysis of the dynamic behavior is carried out by means of an analytical solution approximated by the averaging method and by numerical computational simulations. The absorber reduces the phenomena caused by the Sommerfeld effect and reduces the region of instability.
  • ItemEditorial
    Editorial for Special Issue Dedicated to Vibrations and Control of Elastic Structures (VCES)
    (2022-09-01) Balthazar, José Manoel [UNESP]; Universidade Estadual Paulista (UNESP); Federal University of Technology at Ponta Grossa
  • ItemTrabalho apresentado em evento
    A Self-protecting Approach for Service-oriented Mobile Applications
    (2021-01-01) Martins, Ronaldo Rodrigues [UNESP]; de Oliveira Camargo, Marcos Paulo [UNESP]; Passini, William Filisbino [UNESP]; Campos, Gabriel Nagassaki [UNESP]; Affonso, Frank José [UNESP]; Universidade Estadual Paulista (UNESP)
    The evolution of software systems in the last 10 years has brought new challenges for the development area, especially for service-oriented Mobile Applications (MobApps). In the mobile computing domain, the integration of MobApps into service-based systems has been a feasible alternative to boost the capacity of processing and storage of such applications. In parallel, this type of application needs monitoring approaches mainly due to the need of dealing with a large number of users, continuous changes in the execution environment, and security threats. Besides that, most MobApps do not present the self-protecting property by default, resulting in a number of adverse situations, such as integrity of execution, reliability, security, and adaptations at runtime. The principal contribution of this paper is an approach based on MAPE-K (Monitor-Analyze-Plan-Execute over Knowledge) loop and machine learning techniques to ensure self-protecting features in MobApps, in particular, those based on services. Experimental results showed that this approach can autonomously and dynamically mitigate threats, making these applications more trustworthy and intrusion-safe. Our approach has good potential to contribute to the development of MobApps, going beyond existing approaches.
  • ItemTrabalho apresentado em evento
    A Systematic Mapping on the Use of Data Mining for the Face-to-Face School Dropout Problem
    (2021-01-01) Rondado de Sousa, Leandro [UNESP]; Oliveira de Carvalho, Veronica [UNESP]; Penteado, Bruno Elias; Affonso, Frank José [UNESP]; Universidade Estadual Paulista (UNESP); Universidade de São Paulo (USP)
    Dropout is a critical problem that affects institutions worldwide. Data mining is an analytical solution that has been used to deal with it. Typically, data mining follows a structured process containing the following general steps: data collection, pre-processing, pattern extraction, post-processing (validation). Until know, it is not known how data mining has been used to address the dropout problem in face-to-face education considering all steps of the process. For that, a Systematic Literature Mapping was conducted to identify and analyze the primary studies available in the literature to address some research questions. The aim was to provide an overview of the aspects related to data mining steps in the presented context, without going into details about specific techniques, but about the solutions themselves (for example, imbalanced techniques, instead of SMOTE). 118 papers were selected considering a period of 10 years (01/01/2010 to 31/12/2020).
  • ItemTrabalho apresentado em evento
    MODE SATURATION, MODE COUPLING AND ENERGY HARVESTING FROM AMBIENT VIBRATION IN A PORTAL FRAME STRUCTURE
    (2014-01-01) Balthazar, J. M. [UNESP]; Rocha, R. T. [UNESP]; Brasil, R. M.F.L.; Tusset, A. M.; de Pontes, B. R. [UNESP]; Silveira, M. [UNESP]; Universidade Estadual Paulista (UNESP); Universidade Federal do ABC (UFABC); Parana Federal Technological University
    There has been much recent interest in the concepts of electro-mechanical systems that are able to scavenge, or harvest energy from their operating environment. Here, we present the extraction of energy from a simple portal frame structure excited via its second (first symmetric) mode. As 2:1 internal resonance is present between that mode and the first (sway) mode, the phenomenon of mode saturation and energy exchange (modal coupling) occurs. Energy pumped into the system through the second (vertical) mode is partially transferred to the horizontal (sway) mode. This paper presents results of numerical simulations of these phenomena and energy harvesting using a nonlinear piezoelectric material as a means of energy transduction. An evaluation of the energy available for harvesting in each of the considered modes is computed.
  • ItemTrabalho apresentado em evento
    VIBRATIONS DUE TO IMPACT IN A NON IDEAL MECHANICAL SYSTEM WITH A NON-LINEAR HERTZIAN CONTACT MODEL
    (2014-01-01) Navarro, Helio A.; Balthazar, Jose M.; Brasil, Reyolando M.L.R.F.; Universidade de São Paulo (USP); Federal University of ABC
    This work analyses the post impact behavior of a mechanical system consisting of an oscillator and an unbalanced non-ideal electrical motor. The impact between the mechanical system and a rigid wall is based on the assumption that the impacting bodies undergo local deformations. The method used in the present work is similar to the Discrete Element Method for particle systems modeled with a soft-sphere mechanism. The contact forces are modeled using a nonlinear damped Hertzian Spring-Dashpot system. The mathematical model of the mechanical system is represented by a set of nonlinear ordinary differential equations. The transient and steady-state responses are discussed. As the motor is considered a non ideal energy source, the Sommerfeld effect is also analyzed. The impact model is first applied for a single freely falling particle and then in the proposed mechanical system. Non-dimensional expressions for the contact force and numerical simulations of the mechanical system behavior are also presented.
  • ItemArtigo
    Chaotic saddles and interior crises in a dissipative nontwist system
    (2023-02-01) Simile Baroni, R. [UNESP]; De Carvalho, R. Egydio [UNESP]; Caldas, I. L.; Viana, R. L.; Morrison, P. J.; Universidade Estadual Paulista (UNESP); Universidade de São Paulo (USP); Universidade Federal do Paraná (UFPR); University of Texas at Austin
    We consider a dissipative version of the standard nontwist map. Nontwist systems present a robust transport barrier, called the shearless curve, that becomes the shearless attractor when dissipation is introduced. This attractor can be regular or chaotic depending on the control parameters. Chaotic attractors can undergo sudden and qualitative changes as a parameter is varied. These changes are called crises, and at an interior crisis the attractor suddenly expands. Chaotic saddles are nonattracting chaotic sets that play a fundamental role in the dynamics of nonlinear systems; they are responsible for chaotic transients, fractal basin boundaries, and chaotic scattering, and they mediate interior crises. In this work we discuss the creation of chaotic saddles in a dissipative nontwist system and the interior crises they generate. We show how the presence of two saddles increases the transient times and we analyze the phenomenon of crisis induced intermittency.
  • ItemArtigo
    Lagrangian descriptor and escape time as tools to investigate the dynamics of laser-driven polar molecules
    (2023-02-01) Forlevesi, M. D. [UNESP]; Egydio De Carvalho, R. [UNESP]; De Lima, Emanuel F.; Universidade Estadual Paulista (UNESP); Universidade Federal de São Carlos (UFSCar)
    We consider the nonlinear dynamics of a diatomic polar molecule under a linearly polarized laser field. We assume a model in which the molecule dipole is coupled with a time-dependent electric field. This system presents a bound energy region where the atoms are bound, and a free-energy region where the atoms are dissociated. Due to the nonalignment between the dipole axis and the laser direction, and the time dependence of the external field, this system presents two and a half degrees of freedom, namely the vibrational degree, the rotation degree, and the time. To investigate the system dynamics, instead of using the Poincaré surface-of-section technique, we propose the use of the Lagrangian descriptor associated with the escape times. The Lagrangian descriptor is a quantity that reveals complex structures in the phase space, whereas the escape times are the time span in which a trajectory is initially in the bound region before escaping to the unbound region. The combination of these two quantities allows us to distinguish between real stability regions from other complex structures, including stickiness regions, and a different formation, which we call escape islands. With the help of these tools, we find that for high-field amplitudes the inclusion of rotation leads to an increase of the stability regions, which implies a decrease of the dissociation in comparison with the one-dimensional case.
  • ItemTrabalho apresentado em evento
    MODELING AND ANALYSIS OF A BALL AND BEAM SYSTEM INCLUDING IMPACTS AND DRY FRICTION
    (2014-01-01) Colón, Diego; Bueno, Átila Madureira [UNESP]; Diniz, Ivando S. [UNESP]; Balthazar, Jose M. [UNESP]; Universidade de São Paulo (USP); Universidade Estadual Paulista (UNESP)
    The Ball and Beam system is a common didactical plant that presents a complex nonlinear dynamics. This comes from the fact that the ball rolls over the beam, which rotates around its barycenter. In order to deduce the system's equations, composition of movement must be applied, using a non-inertial reference frame attached to the beam. In the Literature, a common hypothesis is to suppose that the ball rolls without slipping. If a viscous friction is supposed to be present, a simpler situation is obtained, where Lagrangean mechanics can be applied, and no contact force is known. Even then, the dynamics is very nonlinear. However, this model does not include all the relevant phenomena, such as ball's slipping at higher beam's inclination angles, dry friction between the ball and the beam, and impacts between: 1) the ball and the ends of the beam, and 2) the beam and the base (ground). These additions to the model impose the necessity to calculate, in a simulation setting, the contact forces, and the Newton's approach to determine the system's equations becomes more convenient. Also, discontinuities in the model are introduced, and the simpler mathematical object for model such systems are the differential inclusion systems. In this work, we deduce the Ball and Beam differential inclusion system, including dry friction and the impact between the ball and beam. We also present simulation results for the corresponding differential inclusion system in a typical situation.
  • ItemTrabalho apresentado em evento
    THE USE OF WAVELETS ANALYSIS TO CHARACTERIZE THE DYNAMIC BEHAVIOR OF ENERGY TRANSFER VIBRATIONAL SYSTEMS
    (2014-01-01) Iliuk, Itamar; Balthazar, José M. [UNESP]; Tusset, Angelo M.; Piccirillo, Vinicius; Brasil, Reyolando M.L.R.F.; Piqueira, José R.C.; Universidade de São Paulo (USP); Universidade Estadual Paulista (UNESP); UTFPR; Modeling and Applied Social Sciences
    This paper describes the use of wavelet analysis for identification of regular and irregular behavior of dynamical systems. We are focused in single and double-well potential energy harvesting systems that present either periodic or chaotic behavior. To identify the behavior of dynamical systems is of major importance in predicting possible energy harvesting from that system. Using Morlet wavelets, the oscillatory motions of a set of systems were identified with good accuracy. The visualization of the scalograms and global energy spectrum are very useful tools to validate the type of motion found, periodic, quasi-periodic or chaotic. Wavelet analysis can be used to find which amplitude and frequency of operation that generates more energy for each model. Wavelet analysis is a technique used as a tool to assist the validation of the presence of chaos in dynamic systems, together with well consolidated techniques as Lyapunov exponents and Poincare maps.
  • ItemTrabalho apresentado em evento
    A NOVEL RANK CORRELATION MEASURE FOR MANIFOLD LEARNING ON IMAGE RETRIEVAL AND PERSON RE-ID
    (2022-01-01) Valem, Lucas Pascotti [UNESP]; Kawai, Vinicius Atsushi Sato [UNESP]; Pereira-Ferrero, Vanessa Helena [UNESP]; Pedronette, Daniel Carlos Guimarães [UNESP]; Universidade Estadual Paulista (UNESP)
    Effectively measuring similarity among data samples represented as points in high-dimensional spaces remains a major challenge in retrieval, machine learning, and computer vision. In these scenarios, unsupervised manifold learning techniques grounded on rank information have been demonstrated to be a promising solution. However, various methods rely on rank correlation measures, which often depend on a proper definition of neighborhood size. On current approaches, this definition may lead to a reduction in the final desired effectiveness. In this work, a novel rank correlation measure robust to such variations is proposed for manifold learning approaches. The proposed measure is suitable for diverse scenarios and is validated on a Manifold Learning Algorithm based on Correlation Graph (CG). The experimental evaluation considered 6 datasets on general image retrieval and person Re-ID, achieving results superior to most state-of-the-art methods.
  • ItemTrabalho apresentado em evento
    GRAPH CONVOLUTIONAL NETWORKS AND MANIFOLD RANKING FOR MULTIMODAL VIDEO RETRIEVAL
    (2022-01-01) de Almeida, Lucas Barbosa [UNESP]; Valem, Lucas Pascotti [UNESP]; Pedronette, Daniel Carlos Guimarães [UNESP]; Universidade Estadual Paulista (UNESP)
    Despite the impressive advances obtained by supervised deep learning approaches on retrieval and classification tasks, how to acquire labeled data for training remains a challenging bottleneck. In this scenario, the need for developing more effective content-based retrieval approaches capable of taking advantage of multimodal information and advances in unsupervised learning becomes imperative. Based on such observations, we propose two novel approaches that combine Graph Convolutional Networks (GCNs) with rank-based manifold learning methods. The GCN models were trained in an unsupervised way, using the Deep Graph Infomax algorithm, and the proposed approaches employ recent rank-based manifold learning methods. Multimodal information is exploited through pre-trained CNNs via transfer learning for extracting audio, image, and video features. The proposed approaches were evaluated on three public action recognition datasets. High-effective results were obtained, reaching relative gains up to +29.44% of MAP compared to baseline approaches without GCNs. The experimental evaluation also considered classical and recent baselines in the literature.
  • ItemArtigo
    On the impact of mode transition on phased transactional memory performance
    (2023-03-01) Munoz Morales, Catalina; Honorio, Bruno; de Carvalho, Joao P.L.; Baldassin, Alexandro [UNESP]; Araujo, Guido; Universidade Estadual de Campinas (UNICAMP); Universidade Estadual Paulista (UNESP); University of Alberta
    Several transactional memory implementations that employ state-of-the-art software and hardware techniques to deliver performance have been investigated in the last decade. Phased-based Transactional Memory (PhTM) systems run transactions in phases, such that all transactions in a phase execute in the same (hardware/software) mode. In PhTM, a runtime monitors the execution and decides when to change all transactions to another execution mode. Identifying the right moment to perform a mode transition is a central problem to achieve performance in PhTM systems. This article analyzes PhTM and provides a characterization of mode transitions and their impact on performance. We consider three PhTM implementations: (i) PhTM*, the first phased-based TM designed; (ii) Commit Throughput Measurement (CTM), a general-purpose runtime; and (iii) GoTM, a Graph-oriented runtime. We conduct a performance analysis to identify the drawbacks and benefits of each PhTM implementation with respect to their associated parameters. Results with speedups of up to 10× over the sequential baseline for CTM show that this mechanism generally shows better performance for a diverse set of applications.
  • ItemTrabalho apresentado em evento
    Performance Comparison of Speculative Taskloop and OpenMP-for-Loop Thread-Level Speculation on Hardware Transactional Memory
    (2022-01-01) Salamanca, Juan [UNESP]; Universidade Estadual Paulista (UNESP)
    Speculative Taskloop (STL) is a loop parallelization technique that takes the best of Task-based Parallelism and Thread-Level Speculation to speed up loops with may loop-carried dependencies that were previously difficult for compilers to parallelize. Previous studies show the efficiency of STL when implemented using Hardware Transactional Memory and the advantages it offers compared to a typical DOACROSS technique such as OpenMP ordered. This paper presents a performance comparison between STL and a previously proposed technique that implements Thread-Level Speculation (TLS) in the for worksharing construct (FOR-TLS) over a set of loops from cbench and SPEC2006 benchmarks. The results show interesting insights on how each technique can be more appropriate depending on the characteristics of the evaluated loop. Experimental results reveal that by implementing both techniques on top of HTM, speed-ups of up to 2.41× can be obtained for STL and up to 2× for FOR-TLS.
  • ItemArtigo
    Feature augmentation based on manifold ranking and LSTM for image classification[Formula presented]
    (2023-03-01) Pereira-Ferrero, Vanessa Helena [UNESP]; Valem, Lucas Pascotti [UNESP]; Pedronette, Daniel Carlos Guimarães [UNESP]; Universidade Estadual Paulista (UNESP)
    Image classification is a critical topic due to its wide application and several challenges associated. Despite the huge progress made last decades, there is still a demand for context-aware image representation approaches capable of taking into the dataset manifold for improving classification accuracy. In this work, a representation learning approach is proposed, based on a novel feature augmentation strategy. The proposed method aims to exploit available contextual similarity information through rank-based manifold learning used to define and assign weights to samples used in augmentation. The approach is validated using CNN-based features and LSTM models to achieve even higher accuracy results on image classification tasks. Experimental results show that the feature augmentation strategy can indeed improve the accuracy of results on widely used image datasets (CIFAR10, Stanford Dogs, Linnaeus5, Flowers102 and Flowers17) in different CNNs (ResNet152, VGG16, DPN92). The results indicate gains up to 20% and show the potential of the developed approach in achieving higher accuracy results for image classification.
  • ItemArtigo
    Evaluation of different spatial techniques to analyze the occurrence of dengue in Rio Claro - SP.
    (2019-01-01) Domingues, Jacqueline [UNESP]; Govone, José Sílvio [UNESP]; Universidade Estadual Paulista (UNESP)
    In this paper we have developed a study of dengue fever in Rio Claro, Sao Paulo - Brazil. The Municipal Health Foundation of Rio Claro provided data about reported cases of dengue fever in 2011. The main objective was to analyze both the spatial distribution of the disease in the city, by Census tracts, and the relationship of the disease with socioeconomic factors. Two types of spatial models were applied to the data: the SAR model Simultaneous Autoregressive Models and SEM - Simultaneous Error Models. We also fitted the classic linear model, just for comparison to the two spatial models. The results showed that dengue is related to socioeconomic factors and, through the models it was possible to identify which one was statistically significant. Thematic maps have identified the areas that have the highest concentration of the disease.
  • ItemArtigo
    Estudo comparativo do indice de GINI de Minas Gerais entre os anos de 1990 a 2010 via Krigagem Ordinaria e Krigagem Particao
    (Univ Federal Alfenas, 2023-01-01) Abreu, Reinaldo Junio Dias de; Almeida, Maria de Fatima F.; Duraes, Gustavo Rafael Nunes; Govone, Jose Silvio [UNESP]; Inst Fed Norte Minas Gerais IFNMG; Universidade Estadual Paulista (UNESP)
    The GINI index is a measure of inequality widely used to verify the income distribution of a region, it indicates the difference in income from the poorest to the richest, and varies numerically between values from 0 to 1, where the highest value indicates large concentration of income. Performing the analysis of this index in the State of Minas Gerais, using data from Atlas IBGE from 1991 to 2010, it was possible to build maps of Ordinary Kriging and Split Kriging using the Spatial Product Partition Model (SPPM). With the Split Kriging, a division in the sampling grid of the state was obtained, which showed the separation of the central region, the most economically developed in the state, in the 3 analysed periods, and the fluctuation of the GINI indices over the periods in the other regions. We conclude that there was a gradual increase in the GINI index in practically most of the mesoregions of the North of Minas, and the West of Minas remained with the lowest index during all periods compared to the other regions.
  • ItemArtigo
    Desenvolvimento de habilidades e invenções robóticas para impactos sociais no contexto de formação em Matemática
    (Programa de Pós-Graduação em Educação para a Ciência, Universidade Estadual Paulista (UNESP), Faculdade de Ciências, campus de Bauru., 2023-05-19) Azevedo, Greiton Toledo De; Maltempi, Marcus Vinicius [UNESP]; Instituto Federal de Educação; Universidade Estadual Paulista (UNESP)
    We aim to identify and understand the process of developing mathematical skills during the production of robotic devices for Parkinson's disease. Based on qualitative methodology, data were collected with high school students, with recorded videos, photographs, field notes, computer programs and interviews. Such data were analyzed in light of the theoretical assumptions of Constructionism and organized into sections that correspond and thematize the inventions developed. The results point to the development of reasoning, argumentation, communication, representation, and engagement among the students. Such mathematical skills are idiosyncratically constituted by the actions and intellectual productions in an environment of mathematical training that favors autonomy and non-hierarchical dialogue. The skills are understood in three dimensions: Technique, Conceptual and Behavioral. Data also show a process characterized by the non-linear and organic dynamics of training, pointing to the rupture of the content-example-exercise triad and the counterproductive bureaucracies of the classroom.
  • ItemTrabalho apresentado em evento
    Guidelines for the Application of Data Mining to the Problem of School Dropout
    (2022-01-01) de Carvalho, Veronica Oliveira [UNESP]; Penteado, Bruno Elias; de Sousa, Leandro Rondado [UNESP]; Affonso, Frank José [UNESP]; Universidade Estadual Paulista (UNESP); Universidade de São Paulo (USP)
    Dropout is a complex phenomenon based on interrelated factors such as personal, institutional, structural, sociocultural, among other ones. It represents a waste of resources for students, their families, schools and society, and continues to be a challenge for educational institutions. In the last decade, the growing amount of data from educational institutions and the emergence of data science have led to data mining methodologies to explore this problem empirically. In this work, we map the literature on how data mining has been addressed face-to-face dropout. We synthesize different aspects, all of them related to steps of a generic data mining process. Our findings reveal a low level of formalism in theories, methodologies and pre-processing steps, with most papers making comparisons of different algorithms and features on the data available in the institution’s information system. Finally, we present some guidelines that can be used to improve the research on this topic.