DAMNETS: A Deep Autoregressive Model for Generating Markovian Network Time Series

Jase Clarkson, Mihai Cucuringu, Andrew Elliott, Gesine Reinert
Proceedings of the First Learning on Graphs Conference, PMLR 198:23:1-23:19, 2022.

Abstract

Generative models for network time series (also known as dynamic graphs) have tremendous potential in fields such as epidemiology, biology and economics, where complex graph-based dynamics are core objects of study. Designing flexible and scalable generative models is a very challenging task due to the high dimensionality of the data, as well as the need to represent temporal dependencies and marginal network structure. Here we introduce DAMNETS, a scalable deep generative model for network time series. DAMNETS outperforms competing methods on all of our measures of sample quality, over both real and synthetic data sets.

Cite this Paper

BibTeX
@InProceedings{pmlr-v198-clarkson22a, title = {DAMNETS: A Deep Autoregressive Model for Generating Markovian Network Time Series}, author = {Clarkson, Jase and Cucuringu, Mihai and Elliott, Andrew and Reinert, Gesine}, booktitle = {Proceedings of the First Learning on Graphs Conference}, pages = {23:1--23:19}, year = {2022}, editor = {Rieck, Bastian and Pascanu, Razvan}, volume = {198}, series = {Proceedings of Machine Learning Research}, month = {09--12 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v198/clarkson22a/clarkson22a.pdf}, url = {https://proceedings.mlr.press/v198/clarkson22a.html}, abstract = {Generative models for network time series (also known as dynamic graphs) have tremendous potential in fields such as epidemiology, biology and economics, where complex graph-based dynamics are core objects of study. Designing flexible and scalable generative models is a very challenging task due to the high dimensionality of the data, as well as the need to represent temporal dependencies and marginal network structure. Here we introduce DAMNETS, a scalable deep generative model for network time series. DAMNETS outperforms competing methods on all of our measures of sample quality, over both real and synthetic data sets. } }
Endnote
%0 Conference Paper %T DAMNETS: A Deep Autoregressive Model for Generating Markovian Network Time Series %A Jase Clarkson %A Mihai Cucuringu %A Andrew Elliott %A Gesine Reinert %B Proceedings of the First Learning on Graphs Conference %C Proceedings of Machine Learning Research %D 2022 %E Bastian Rieck %E Razvan Pascanu %F pmlr-v198-clarkson22a %I PMLR %P 23:1--23:19 %U https://proceedings.mlr.press/v198/clarkson22a.html %V 198 %X Generative models for network time series (also known as dynamic graphs) have tremendous potential in fields such as epidemiology, biology and economics, where complex graph-based dynamics are core objects of study. Designing flexible and scalable generative models is a very challenging task due to the high dimensionality of the data, as well as the need to represent temporal dependencies and marginal network structure. Here we introduce DAMNETS, a scalable deep generative model for network time series. DAMNETS outperforms competing methods on all of our measures of sample quality, over both real and synthetic data sets.
APA
Clarkson, J., Cucuringu, M., Elliott, A. & Reinert, G.. (2022). DAMNETS: A Deep Autoregressive Model for Generating Markovian Network Time Series. Proceedings of the First Learning on Graphs Conference, in Proceedings of Machine Learning Research 198:23:1-23:19 Available from https://proceedings.mlr.press/v198/clarkson22a.html.

Related Material