• Entrar
    Ver item 
    •   Página inicial
    • BIBLIOTECA DIGITAL: Trabalhos de Especialização
    • Ciências Exatas e da Terra
    • Data Science & Big Data
    • Ver item
    •   Página inicial
    • BIBLIOTECA DIGITAL: Trabalhos de Especialização
    • Ciências Exatas e da Terra
    • Data Science & Big Data
    • Ver item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Modelos preditivos para Doença Cardiovacular

    Thumbnail
    Visualizar/Abrir
    R - E - EDSON LUIZ DOS SANTOS.pdf (633.7Kb)
    Data
    2022
    Autor
    Santos, Edson Luiz dos
    Metadata
    Mostrar registro completo
    Resumo
    Resumo: Este estudo foi realizado com o intuito de auxiliar a identificação antecipada de pessoas que tenham tendências de contrair doença cardiovascular.A base de dados utilizada neste estudo, é uma base pública disponível no kaggle, ela possui 918 registros com 11 variáveis explicativas e 1 variável resposta. A primeira parte do processo, foi efetuar uma análise exploratória das variáveis e identificar quais possuem maior influência na causa de doenças cardiovasculares, e quais variáveis podiam ajudar a prever quais pacientes podem desenvolver doença cardiovascular, além de identificar quais são os principais causadores deste tipo de doença. A segunda parte do processo, para responder os fatores da pesquisa, foram utilizadas as metodologias, modelo logístico, arvore de decisão e random forest, estas metodologias foram escolhidas por possuírem características de modelos binários, ou seja, modelos de classificação. A terceira parte do processo, foi utilizada para testar a acurácia das técnicas aplicadas, os modelos são capazes de realizar a predição da doença, conforme as características do conjunto de dados. Todos os testes resultaram em uma excelente assertividade, o modelo logístico obteve a melhor performance, atingiu aproximadamente 87%, por se tratar de um modelo mais simples e a melhor acurácia, o modelo logístico foi o escolhido para auxiliar na predição deste tipo de doença.
     
    Abstract: This study was carried out with the aim of helping the early identification of people who have a tendency to contract heart disease. The database used in this study is a public database available on kaggle, there are 918 records with 11 explanatory variables and 1 response variable. The first part of the process was to carry out an exploratory analysis of the variables and identify which ones have the greatest influence on the cause of heart disease, and which variables could help predict which patients may develop heart disease, in addition to identifying which are the main causes of this type of disease. The second part of the process, to answer the research factors, were used the methodologies, logistic model, decision tree and random forest, these methodologies were chosen because they have characteristics of classification models. The third part of the process was used to test the accuracy of the applied techniques, the models are able to predict the disease, according to the characteristics of the data set. All tests resulted in excellent assertiveness, the logistic model obtained the best performance, reaching approximately 87%, because it is a simpler model and the best accuracy, the logistic model was chosen to assist in the prediction of this type of disease.
     
    URI
    https://hdl.handle.net/1884/79919
    Collections
    • Data Science & Big Data [138]

    DSpace software copyright © 2002-2022  LYRASIS
    Entre em contato | Deixe sua opinião
    Theme by 
    Atmire NV
     

     

    Navegar

    Todo o repositórioComunidades e ColeçõesPor data do documentoAutoresTítulosAssuntosTipoEsta coleçãoPor data do documentoAutoresTítulosAssuntosTipo

    Minha conta

    EntrarCadastro

    Estatística

    Ver as estatísticas de uso

    DSpace software copyright © 2002-2022  LYRASIS
    Entre em contato | Deixe sua opinião
    Theme by 
    Atmire NV