This project aims to make you discover the field of machine learning through a simple exercise: the automatic detection of diabetes. It is therefore dedicated to beginners and to anyone who wants to discover this field. As this project is an introduction, you will find here only simple concepts and the bare minimum. So if you are already experienced in the field, don't hesitate to take a look at our other projects.
This project can be perform on Windows, MacOS or Linux.Moreover, no GPU card is required.
Concerning python you don't need to be an expert of this language but a minimal programming background is required.
This project is composed of 3 folders and a Medium article (here) containing our approach and additional explanations on the field.
- The dataset folder:
contains the data set "Pima_Indians_Diabetes" from kaggle - The machine learning folder :
Contains a code treating the problem from a machine learning point of view as well as a file containing the required facilities. - The deep learning folder:
Contains a code treating the problem from a deep learning point of view as well as a file containing the required facilities.
Each of these lines corresponds to a patient with or without diabetes.
The column "Outome" corresponds to the labels, so it indicates if the patient has diabetes `valued 1` or not `valued 0`. The 8 other features : [Pregnancies, Glucose, BloodPressure, SkinThickness, Insulin, BMI, DiabetesPedigreeFunction, Age] are used as descriptionn to help make a prediction. Choose your solving method "Machine learning" or "Deep learning" and install the necessary dependencies :
cd MachineLearning
pip install -r requirement.txt
cd DeepLearning
pip install -r requirement.txt
# Éditez Ligne 24
df = pd.read_csv(github_path)
# Exécuter le code
python simpleDetection.py
Kaggle : https://www.kaggle.com
