APPLYING DEEP-CLUSTERING APPROACHES TO CREATE MEANINGFUL REPRESENTATIONS OF ECG BEAT MORPHOLOGY

Introduction

This is a PyTorch implementation of our Deep Learning on Computational Accelerators course project detailed in Report.pdf

by:

Name	E-mail
Amit Rotner	[email protected]
Shaked Doron	[email protected]

Installation

1. Clone this repository.
2. Install and activate the environment using the following commands:

conda env create -f environment.yml
conda activate final_project

Project Structure

.
├── cs236781/                                       # Helper files
│   ├── plot.py                                     # Helper function to plot experiments results 
│   └── unit.py                                     # Helper classes to represent the result of fitting a model
├── data/                                           # Dataset directory
├── ptb.py                                          # The implementation of the CNN network for the PTB dataset
├── mit_bih.py                                      # The implementation of the CNN network for the MIT-BIH dataset
├── autoencoder.py                                  # The implementation of the Autoencoder network
├── clustering.py                                   # The implementation of the clustering layer, kmeans, target distribution, and clustering predictions calculation 
├── utils.py                                        # Helper class implementing metrics for clustering evaluation
├── training.py                                     # The implementation of the model training and testing functions
├── PTB classification with CNN.ipynb               # A Jupyter notebook to perform and display PTB classification using the CNN network
├── MIT-BIH classification with CNN.ipynb           # A Jupyter notebook to perform and display MIT-BIH classification using the CNN network
├── DCEC ptb.ipynb                                  # A Jupyter notebook to perform and display PTB classification using the DCEC network
├── DCEC mit-bih.ipynb                              # A Jupyter notebook to perform and display MIT-BIH classification using the DCEC network
├── Experiments.ipynb                               # A Jupyter notebook to perform DCEC experiments varying on d
├── Report.pdf                                      # Project report
├── environment.yml
├── LICENSE
└── README.md

Reproducing results

1. Download datasets from: https://drive.google.com/drive/folders/1fefvwQfyTafnq0rybXCWT9wmSElzP58A?usp=sharing and place it in ./data/ folder.
2. Run the relevant Jupyter notebook.