NCovid ML Modules is a standalone library for machine learning applications, compatible with receiving requests as web-point.
The NCovid-ml-modules design goal is improving formatting and facilitating to make machine learning pipelines readable and optimized. The main idea is to optimize the time-series data manipulation by preprocessed workflow and submit to the use of many applications to data regression problems. The library is based on OOP and multiple goals could be done such as preprocess data, models grid search, and evaluation.
NCovid ML Modules uses a number of open source projects to work properly:
- Python 3 - a powerful general-purpose programming language
- Pandas - fast, powerful, flexible and easy to use open source data analysis and manipulation tool
- NumPy - comprehensive mathematical functions
- Keras - an open-source software library that provides a Python interface for artificial neural networks
- Matplotlib - plotting library
And of course NCovid ML Modules itself is open source with a public repository on GitHub.
This code runs on Python 3.7
We have tested the library in Ubuntu 20.04, 19.04, 18.04, and 16.04, but it should be easy to compile on other platforms.
Please make sure that it has installed all the required dependencies. A list of items to be installed using pip install can be running as following:
pip install -r requirements.txt
Project Folder Structure and Files
- src : main Python package with source of the model.
- docs : contains documentation of the project.
- jupyter-notebook : contains jupyter notebooks evaluation and modeling experimentation.
People involved in this project
| Role | Responsibility | Full name | orcid |
|---|---|---|---|
| Data Scientist | Scrum Master | Davi Santos | ORCID |
| Data Scientist | Tech Leader | Dunfrey P. Aragão | ORCID |
| Data Scientist | Developer | Emerson Vilar | ORCID |
MIT
Free Software, Hell Yeah!