项目作者: ldaniel

项目描述 :
Practicing Advanced Predictive Analytics in Python.
高级语言: Jupyter Notebook
项目地址: git://github.com/ldaniel/fgv-advanced-predictive-analytics.git
创建时间: 2020-03-08T15:32:28Z
项目社区:https://github.com/ldaniel/fgv-advanced-predictive-analytics

开源协议:Apache License 2.0

下载


Introduction

In the predictive analysis discipline we aim to develop statistical models, based on data, for a particular outcome of interest. From this model, we make sure that behavioral learning and past experiences have a model with good generalization.

In this project, we developed a set of tasks to go further in Predictive Analytics by exploring some Deep Learning techniques.

This website intends to present the work analysis for the “Análise Preditiva Avançada” class assignment.

  1. Trabalho em grupo de 3 a 4 alunos envolvendo técnicas de aprendizado supervisionado
  2. de máquina com Deep learning.
  3. Possíveis trabalhos:
  4. - Classificação ou Previsão, Dados Numéricos ou Categóricos, Estilo de problema
  5. similar às Regressões Logística ou Linear;
  6. - Principalmente no caso de processamento de texto em linguagem natural;
  7. - Classificação de imagens com Redes Convolutivas;
  8. - Previsão de Texto com Redes Sequenciais;
  9. - Geração de conteúdo (música por exemplo) com redes Auto-Generativas;
  10. - Mix de estilos artísticos com Redes Convolutivas / Auto-Generativas.
  11. Material a ser entregue:
  12. Se fizer em Python (recomendado): Jupyter Notebook com base e resultados (no caso de
  13. dados não submetidos à confidencialidade) ou apenas o Jupyter Notebook (com algumas
  14. referências de dados que possam "validar" o modelo de rede neural entregue).
  15. Se fizer em R: Entregar a base de dados e o Rmarkdown. As bibliotecas de machine
  16. learning mais utilizadas (scikit-learn, tensorflow e keras) estão disponíveis também
  17. no R, porém com um print-end. É necessário instalar o Python para executalas.

:link: See the final website report in Rodrigo Gonçalves’ Kaggle profile at: https://www.kaggle.com/rodrigonca/advanced-predictive-analysis-cnn-implementation.

:octocat: Alternatively, run a binder container:

Binder

Professor

  • Gustavo Mirapalheta

Authors / Students

Profile Name E-mail
Daniel Campos daniel.ferraz.campos@gmail.com"">(daniel.ferraz.campos@gmail.com)
Leandro Daniel (contato@leandrodaniel.com)
Rodrigo Goncalves rodrigo.goncalves@me.com"">(rodrigo.goncalves@me.com)
Ygor Lima ygor_redesocial@hotmail.com"">(ygor_redesocial@hotmail.com)

Project Organization


  1. ├── LICENSE
  2. ├── Makefile <- Makefile with commands like `make data` or `make train`
  3. ├── README.md <- The top-level README for developers using this project.
  4. ├── data
  5. ├── external <- Data from third party sources.
  6. ├── interim <- Intermediate data that has been transformed.
  7. ├── processed <- The final, canonical data sets for modeling.
  8. └── raw <- The original, immutable data dump.
  9. ├── docs <- A default Sphinx project; see sphinx-doc.org for details
  10. ├── models <- Trained and serialized models, model predictions, or model summaries
  11. ├── notebooks <- Jupyter notebooks.
  12. └── external_examples <- Other interesting notebooks
  13. └── fgv_assignment <- The final class assignment given by Professor Mirapalheta
  14. └── fgv_classes <- All notebooks given by Professor Mirapalheta and Professor Hithoshi
  15. in theirs respective classes
  16. ├── references <- Data dictionaries, manuals, and all other explanatory materials.
  17. ├── reports <- Generated analysis as HTML, PDF, LaTeX, etc.
  18. └── figures <- Generated graphics and figures to be used in reporting
  19. ├── requirements.txt <- The requirements file for reproducing the analysis environment, e.g.
  20. generated with `pip freeze > requirements.txt`
  21. ├── setup.py <- makes project pip installable (pip install -e .) so src can be imported
  22. ├── src <- Source code for use in this project.
  23. ├── __init__.py <- Makes src a Python module
  24. ├── data <- Scripts to download or generate data
  25. └── make_dataset.py
  26. ├── exercises <- Scripts for FGV's class assignments
  27. │ │ └── __init__.py
  28. │ │ └── playground.py
  29. │ │
  30. │ ├── features <- Scripts to turn raw data into features for modeling
  31. │ │ └── build_features.py
  32. │ │
  33. │ ├── models <- Scripts to train models and then use trained models to make
  34. │ │ │ predictions
  35. │ │ ├── predict_model.py
  36. │ │ └── train_model.py
  37. │ │
  38. │ └── visualization <- Scripts to create exploratory and results oriented visualizations
  39. │ └── visualize.py
  40. └── tox.ini <- tox file with settings for running tox; see tox.testrun.org

Project based on the cookiecutter data science project template. #cookiecutterdatascience