Machine Learning Diabetes Prediction - HackoonSpace 2020/2

Conception

This is a Data Science and Machine Learning project that tries to predict early diabetes type two using simple classification algorithms, such as Decision Tree, KNN and SVC. This project was done for study purposes and for an extension activity - HackoonSpace - of my university.

Prerequisites and Resources

The implementation uses Python language with Google Colaboratory and libraries as NumPy, Matplotlib, Pandas and Scikit Learn. Also, the models are trained from the following data set:

1. "Early stage diabetes risk prediction dataset", downloadable at UCI Repository

Steps

• Data collection and exploratory analysis
• Data pre-processing, coding and regularization
• Models training
• Evaluation and validation

Installation

1. Download the repository
2. Open and run the .ipynb files on Google Colaboratory or Jupyter Notebook
   *you can do tests and change some hyperparameters of the models to try better results

Author

• Matheus Vargas Volpon Berto (MathVolps)

References:

• Numpy Reference Guide
• Mlxtend: User Guide Index
• Plotly Python Open Source Graphing Library
• Scikit Learn: Machine Learning in Python
• Seaborn: Statistical Data Visualization.
• Machine Learning A-Z™: Hands-On Python & R In Data Science