Machine Learning Project 1 - CVD Prediction

Authors :

• Anas Himmi ([email protected])
• Zeineb Mellouli ([email protected])
• Antoine Cornaz ([email protected])

Description :

The objective of this project is to develop a machine learning model capable of predicting the presence of cardiovascular disease (CVD) in patients based on their medical information. We explore and compare different algorithms to determine the best approach for this classification task.

Dependencies :

Before running the solution, make sure to have the following Python library installed:

• numpy

You can install them using the following command:

pip install numpy

For visualization purposes in the preprocessing and feature selection phases, we also use matplotlib, seaborn but they are not needed to run the models and the final training.

How to Run the Project (Minimal) :

Run the command:

python run.py

The train and test csv files should be located in a folder named "data" one directory above the location of the run.py file. The predictions will be saved in a file named 'final_submission.csv'

Code Organization :

• implementations.pycontains the mandatory functions for the project

• run.ipynb contains all the code and steps we followed to implement the project. It is the most complete and we invite the reader to follow the steps we took to implement the project in this file.

• run.pycontains the minimal code to generate the submission (might take 3 minutes of loading the data and 3 minutes of training)

• helpers.py contains various helper functions, including data loading, standardization, and utility functions for handling CSV submissions and batch iterations.