Portal de Periódicos da Univali (Universidade do Vale do Itajaí)
Not a member yet
9513 research outputs found
Sort by
Machine Learning-based Method to Label Signals from People with Neurological Injuries
Get PDFABSTRACTInterfaces that use sEMG signals face the challenge of correctly identifyingthe signal while distinguishing it from noise or interference.Although classical techniques like visual inspection and machinelearning methods exist, most studies focus on signals from healthyindividuals. There is a lack of data and methods suitable for signalsfrom individuals with neurological conditions, such as cerebralpalsy and post-stroke. This study analyzes sEMG data from individualswith neurological injuries, using machine learning methods toidentify muscle contractions and rest without pre-processing. Thedata were acquired from people with neurological diseases, such ascerebral palsy and post-stroke. They were extracted using sEMGfrom triceps brachii and extensor carpi radialis muscles. The signalswere not preprocessed and were input as segmented time windowsto three proposed classifiers: Support Vector Machine, Random Forestand an Ensemble Voting classifier. All three classifiers reachedaround 99% accuracy and F1-Score on typical sEMG data, but theresults on abnormal data were inconclusive
Conjunto de Dados para Envenenamento de Modelos Baseados em Self-Training em Fluxo de Dados
Get PDFAbstractIn problems with a large volume of unlabeled data, semi-supervisedlearning techniques, such as self-training, are attractive becausethey make full use of the data and do not require extensive labelingof the data, since it is an expensive process. However, using pseudolabelsto train a model indiscriminately can lead to undue changes inthe model’s decision boundary, which can happen unintentionallyor intentionally, such as in malware classification, where attackerswant to classify malicious software as benign. In this paper, wepropose a dataset for poisoning models based on self-training thatsimulates a data stream, intending to evaluate the robustness ofthese models against intentional or unintentional poisoning byunlabeled instances. Our experiments use models from the MOA-SSframework, and show that models that use incremental trainingand prediction confidence as a criterion for using the unlabeledinstance in training are more susceptible to poisoning
Explorando o Pensamento Computacional nos Anos Iniciais: Uma Jornada de Ensino e Extensão no Ensino Fundamental
Get PDFABSTRACTThis study reports the actions of an extension project that appliedComputational Thinking and robotics concepts in a municipalelementary school. The project, aligned with the CommonNational Curriculum Base (BNCC), aimed to foster logicalreasoning through practical and playful activities. Unpluggedmethodologies were employed, including translating binarynumbers, simulating classification networks, and exploring acomputer maze. These activities introduced students to the coreprinciples of computational thinking, focusing on algorithmicthinking, task decomposition, and problem-solving. The resultsdemonstrated the project's effectiveness, as students engagedactively and developed essential cognitive skills. This experiencehighlights the value of incorporating computational thinking intoearly basic education through playful and adaptable teachingapproaches tailored to diverse school contexts
Automated Extraction of Body Measurements Using Computer Vision
Get PDFAbstractExtracting body measurements manually is a time-consuming andchallenging process that can be impractical without the appropriatetools. This paper proposes a method to obtain body measurements,which can then be used in applications such as finding the perfectclothing fit. By leveraging computer vision, the developed systemcalculates body measurements from a photo submitted by the user,along with their height. Utilizing essential libraries and detaileddigital image processing techniques, we successfully performed thecalculations to return the user’s body measurements. Test resultsdemonstrate that the model achieves high accuracy, underscoringthe efficiency of this approach and offering a swift alternative forobtaining body measurements. The developed method can be employedin numerous other applications that require precise bodymeasurements
O Cenário Atual de Golpes em Redes Sociais: Uma Revisão da Literatura
Get PDFAbstractSocial networks were conceived as platforms for interaction, contentcreation and engagement between users. Over time, browsingthese networks has become an integral part of everyday life, culminatingin billions of profiles created and accessed regularly onvarious platforms. However, the ease of access and limited controlover content by the maintainers have turned social networks into afavorable environment for fraudulent activities, allowing malicioususers to act as scammers in search of illicit benefits. Faced withthis scenario, academia and industry have dedicated themselves toanalyzing and developing methods to identify, detect, prevent andmitigate scams. However, the diversity of scams, many with uniquecharacteristics, requires specific analysis and solutions for eachcase. In this context, this paper proposes a systematization of socialmedia scams addressed in the literature between 2014 and 2024,defining categories and subcategories based on their methodologicalcharacteristics and objectives. The main result is a taxonomymade up of 5 categories and 19 subcategories
Ferramenta para qualificar a Complexidade Cognitiva do código utilizando a técnica Cognitive Driven Development
Get PDFABSTRACTThis work proposed the development of an extension for Visual Studio Code, which supports the PHP programming language, using the Cognitive Driven Development (CDD) method to identify the cognitive complexity of the source code. The extension helps the developer to locate code snippets with unnecessary complexity, encouraging the application of refactoring techniques and good code design practices. For this purpose, research related to CDD was analyzed and the necessary parameters for the creation of a configuration file in JSON were defined. Due to the limitation of the VS Code API, which does not provide the AST (Abstract Syntax Tree) of the code, it was decided to integrate the Php-parser library, written in JavaScript, which proved to be compatible with the Typescript development environment. The tool was validated through tests on a real project, comparing its performance before and after refactoring. In the initial analysis, the extension correctly identified the critical points that needed improvement. After refactoring, a reduction in the identified complexity was observed, evidencing the effectiveness of the tool, even when applied to a small project. In summary, the extension proves to be an effective solution for automating the identification of cognitive complexity, contributing to more efficient software development and continuous maintenance of source code
Fortalecendo a Presença Feminina na Computação: Análise do Impacto do projeto Meninas Digitais Vale do Itajaí no Desenvolvimento Acadêmico
Get PDFABSTRACTDespite increasing career opportunities in computing, womenremain underrepresented in this field due to societal stereotypes andsystemic challenges. The Meninas Digitais Vale do Itajaí project,aims to address gender disparities by promoting initiatives thatencourage women to pursue and persist in computing careers. Thisstudy evaluates the impact of the project's activities on femalestudents in Information Systems and Computer Science at FURBUniversity
Proposta de Metodologia de Análise de Redes Wi-Fi Prediais
Get PDFABSTRACTThe present work proposes a methodology for evaluating building Wi-Fi networks using a spectrum analyzer alongside Wi-Fi network analysis tools. The Agilent CSA Spectrum Analyzer, model N1996A-506, and the tools ManageEngine OpManager and PRTG Professional Wi-Fi Analyzer will be used to measure the Wi-Fi signal of an academic building. The collected Received Signal Strength Indicator (RSSI) data will be imported and modeled in Python code to generate heatmaps that visually represent the quality of the Wi-Fi network coverage. The results obtained from the tools will be compared with those obtained from the analyzer to verify the accuracy and reliability of the reports generated by the tools
Cartilhas de Atividades para Promoção de Igualdade de Gênero e Cultura digital
Get PDFAbstractRecent studies has shown that women have been historically underrepresentedin STEAM. The main goal of this project was todisseminate the contributions of women in Computing whilst promotingComputational Thinking, Digital Culture and Digital Technology– related to the 5th General Competence from BNCC. Inorder to achieve this goal, four booklets were developed to helpto disseminate in a ludic manner the female contributions and thecomputational concepts previously mentioned. This same materialwas presented to educators in workshops and it was also madeavailable in a digital in the project’s website
