Coder Social home page Coder Social logo

luizsci42 / analise-de-sentimentos-pandemia-covid19 Goto Github PK

View Code? Open in Web Editor NEW
0.0 0.0 0.0 9.86 MB

Repositório utilizado para o plano de PIBIC 2020-2021 com o prof. Dr. Hendrik Macedo. Tem como finalidade criar um dataset para treinamento de modelos de aprendizado de máquina sobre as 5 emoções de Ekman e analisar os sentimentos predominantes durante os primeiros 12 meses da pandemia de COVID-19.

Python 10.53% Jupyter Notebook 89.47%
data-science linear-models machine-learning natural-language-processing scikit-learn sentiment-analysis spacy-nlp text-mining text-visualization

analise-de-sentimentos-pandemia-covid19's Introduction

Hi, i'm Luiz Felipe!

Business Intelligence Analyst by Sergipe Parque Tecnológico, allocated at Secretaria de Estado da Fazenda Undergraduate student on Computer Science at Universidade Federal de Sergipe
Researcher at Ludiico Labs

GitHub Luiz Felipe Souza Linkedin: luiz-felipe-souza Docker Hub Kaggle

class LuizFelipe:

    def __init__(self):
        self.name = 'Luiz Felipe Souza'
        self.code = {
            'fields_of_interest': ['Data Science', 'Natural Language Processing'],
            'tools': ['Scikit-learn', 'PowerBI', 'Jupyter Notebook', 'Pentaho', 'Docker'],
            'Backend': ['Python Flask']           
        }


if __name__ == '__main__':
    me = LuizFelipe()
Hi there! I'm Luiz Felipe and I'm cunrrently working as Business Intelligence Analyst on SergipeTec (Sergipe Parque Tecnológico) and undegraduating on Computer Science at Universidade Federal de Sergipe.

I had made some projects at Ludiico Labs, an academic group inside the university, from which i had gained experience on
Android, FullStack Development, Data Science and Natural Language Processing.

In all these projects, I used to be very proactive and criative, collaborating not only with the development, but also
with sugestions, solutions and new strategies when something went wrong.

At my job, I work with processes and tools of Business Intelligence, as ETL (Extract, Transform, Load), reporting with dashboards and training of machine learning models.
      
My interests include Ethical technology development, Data Science, Creative Writing and Inovation.

From @luizsci42

analise-de-sentimentos-pandemia-covid19's People

Contributors

luizsci42 avatar

Watchers

 avatar

analise-de-sentimentos-pandemia-covid19's Issues

Treinar um novo modelo com maior desempenho

Algumas medidas podem ser tomadas para obter um melhor modelo, dentre as quais otimização de parâretros (fine tuning) e/ou testar outro modelo.

Em testes iniciais, consegui obter um weighted f1-score de 58% obtido utilizando cross-validation com 10 folds, a partir de um modelo de logistic regression com parâmetros padrão, sem a necessidade de balancear o conjunto de treinamento. Estou prosseguindo tentando otimizar os parâmetros tanto do LinearSVC, quanto do Logistic Regression.

Outras possíveis medidas também envolvem o pré-processamento do texto, como stemming e lemmatization.

Futuramente, também posso fazer uso do BERT. O seguinte tutorial me parece uma boa introdução: https://www.analyticsvidhya.com/blog/2023/06/step-by-step-bert-implementation-guide/

Revisar dataset de treinamento

Por ter sido montado a partir de um crawler, baseando-se em hashtgs, muitos tweets presentes no dataset podem conter ruído. É necessário verificar as labels e remover o que não for útil.

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.