dynamic_disease_network_ddp

Learning Dynamic and Personalized Comorbidity Networks

Implementation of Learning Dynamic and Personalized Comorbidity Networks from Event Data using Deep Diffusion Processes (AISTATS2020) in Python runnable both on CPU and GPU.

Reference

If you use this code as part of any published research, please acknowledge the following paper:

@article{qian2020learning,
  title={Learning Dynamic and Personalized Comorbidity Networks from Event Data using Deep Diffusion Processes},
  author={Qian, Zhaozhi and Alaa, Ahmed M and Bellot, Alexis and Rashbass, Jem and van der Schaar, Mihaela},
  journal={arXiv preprint arXiv:2001.02585},
  year={2020}
}

Running the Code

Dependency

The minimum dependency requirements are specified in requirements.txt.

You will need an installation of PyTorch (>=1.3) in addition to the standard data science python packages. We highly recommend using the conda distributions.

Project Structure

The model itself is implemented in models.py. The utility functions related to data ingestion and manipulation are implemented in data_loader.py. simulation.py is the entry point to the simulation. To run the simulation, run

python simulation.py

The simulated data set is generated using neurawkes and is located in folder data. The simulation output is a figure sim_result.png. It reproduces the simulation study in the appendix (figure 9).

License

This project is licensed under the MIT License.