EpiRDN: A Learnable Anisotropic Reaction-Diffusion Network for Epidemic Time Series Prediction

Asela Hevapathige
Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337:2137-2151, 2026.

Abstract

Epidemic spread on graphs involves local infection dynamics and directional propagation through population movement. Current epidemic graph neural networks combine these aspects using a single symmetric aggregation operator, which limits the learning of their distinct timescales and fails to capture the asymmetry of real epidemic spread. This leads to node representations collapsing to a common state, losing diversity necessary for distinguishing infection stages. To address these issues, we propose EpiRDN, which discretizes the reaction-diffusion equation on graphs and learns its components end-to-end. EpiRDN features a local reaction network for region-level transitions and an anisotropic diffusion operator that utilizes asymmetric attention for directional infection flow. A feature-conditioned damping coefficient balances preserving local identity and neighborhood aggregation. We demonstrate that feature diversity in EpiRDN decays at most geometrically with depth, providing a significant advantage over existing methods. Experiments on four real-world datasets related to influenza and COVID-19 show consistent improvements, especially at longer forecasting horizons.

Cite this Paper

BibTeX
@InProceedings{pmlr-v337-hevapathige26a, title = {EpiRDN: A Learnable Anisotropic Reaction-Diffusion Network for Epidemic Time Series Prediction}, author = {Hevapathige, Asela}, booktitle = {Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence}, pages = {2137--2151}, year = {2026}, editor = {Perković, Emilija and Malinsky, Daniel}, volume = {337}, series = {Proceedings of Machine Learning Research}, month = {17--21 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v337/main/assets/hevapathige26a/hevapathige26a.pdf}, url = {https://proceedings.mlr.press/v337/hevapathige26a.html}, abstract = {Epidemic spread on graphs involves local infection dynamics and directional propagation through population movement. Current epidemic graph neural networks combine these aspects using a single symmetric aggregation operator, which limits the learning of their distinct timescales and fails to capture the asymmetry of real epidemic spread. This leads to node representations collapsing to a common state, losing diversity necessary for distinguishing infection stages. To address these issues, we propose EpiRDN, which discretizes the reaction-diffusion equation on graphs and learns its components end-to-end. EpiRDN features a local reaction network for region-level transitions and an anisotropic diffusion operator that utilizes asymmetric attention for directional infection flow. A feature-conditioned damping coefficient balances preserving local identity and neighborhood aggregation. We demonstrate that feature diversity in EpiRDN decays at most geometrically with depth, providing a significant advantage over existing methods. Experiments on four real-world datasets related to influenza and COVID-19 show consistent improvements, especially at longer forecasting horizons.} }
Endnote
%0 Conference Paper %T EpiRDN: A Learnable Anisotropic Reaction-Diffusion Network for Epidemic Time Series Prediction %A Asela Hevapathige %B Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2026 %E Emilija Perković %E Daniel Malinsky %F pmlr-v337-hevapathige26a %I PMLR %P 2137--2151 %U https://proceedings.mlr.press/v337/hevapathige26a.html %V 337 %X Epidemic spread on graphs involves local infection dynamics and directional propagation through population movement. Current epidemic graph neural networks combine these aspects using a single symmetric aggregation operator, which limits the learning of their distinct timescales and fails to capture the asymmetry of real epidemic spread. This leads to node representations collapsing to a common state, losing diversity necessary for distinguishing infection stages. To address these issues, we propose EpiRDN, which discretizes the reaction-diffusion equation on graphs and learns its components end-to-end. EpiRDN features a local reaction network for region-level transitions and an anisotropic diffusion operator that utilizes asymmetric attention for directional infection flow. A feature-conditioned damping coefficient balances preserving local identity and neighborhood aggregation. We demonstrate that feature diversity in EpiRDN decays at most geometrically with depth, providing a significant advantage over existing methods. Experiments on four real-world datasets related to influenza and COVID-19 show consistent improvements, especially at longer forecasting horizons.
APA
Hevapathige, A.. (2026). EpiRDN: A Learnable Anisotropic Reaction-Diffusion Network for Epidemic Time Series Prediction. Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 337:2137-2151 Available from https://proceedings.mlr.press/v337/hevapathige26a.html.

Related Material