A GAN-based reduced order model for prediction, data Assimilation and uncertainty quantification

This repository is the official implementation of:

Digital twins based on bidirectional LSTM and GAN for modelling the COVID-19 pandemic (for the Predictive GAN).

Data Assimilation Predictive GAN (DA-PredGAN) Applied to a Spatio-Temporal Compartmental Model in Epidemiology.

Generative model-based framework for parameter estimation and uncertainty quantification applied to a compartmental model in epidemiology (coming soon)

Directories:

• PredGAN (outdated see DA-PredGAN/UQ-PredGAN folder): Prediction using GAN - applied to the spatio-temporal spread of COVID-19 in an idealized town.
• DA-PredGAN: Data assimilation using GAN - applied to the spatio-temporal spread of COVID-19 in an idealized town.
• UQ-PredGAN: Uncertainty quantification using GAN - applied to the spatio-temporal spread of COVID-19 in an idealized town.
• datasets: Datasets of the spatio-temporal spread of COVID-19 in an idealized town.
• GAN_evaluation: New way of evaluating the GAN training.
• Regularization: Regularization to improve the GAN-based Reduced Order Model.
• MCMC: Comparison between the UQ-PredGAN and the Markov chain Monte Carlo (MCMC) methods.

Requirements

To install requirements:

 $ conda env create -f environment.yml 
 $ conda activate py3ml
 $ python -m ipykernel install --user --name=python3 (optional)

Finally, start Jupyter:

 $ jupyter notebook