Welcome aboard BayesAir, a Bayesian framework for modeling and analyzing air transportation network dynamics! 🛫✨
This repository contains a Bayesian model of air traffic network dynamics implemented as a probabilistic program in Pyro.
It also provides our algorithm, CalNF, for learning generative models in data-constrained settings.
This code accompanies our ICLR 2025 paper, Rare event modeling with self-regularized normalizing flows: what can we learn from a single failure? . If you use this code, please cite
@inproceedings{
dawson2025rare,
title={Rare event modeling with self-regularized normalizing flows: what can we learn from a single failure?},
author={Charles Dawson and Van Tran and Max Z. Li and Chuchu Fan},
booktitle={The Thirteenth International Conference on Learning Representations},
year={2025},
url={https://openreview.net/forum?id=gQoBw7sGAu}
}