Graph fission and cross-validation

James Leiner, Aaditya Ramdas
Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, PMLR 238:2638-2646, 2024.

Abstract

We introduce a technique called graph fission which takes in a graph which potentially contains only one observation per node (whose distribution lies in a known class) and produces two (or more) independent graphs with the same node/edge set in a way that splits the original graph’s information amongst them in any desired proportion. Our proposal builds on data fission/thinning, a method that uses external randomization to create independent copies of an unstructured dataset. We extend this idea to the graph setting where there may be latent structure between observations. We demonstrate the utility of this framework via two applications: inference after structural trend estimation on graphs and a model selection procedure we term "graph cross-validation"’.

Cite this Paper

BibTeX
@InProceedings{pmlr-v238-leiner24a, title = { Graph fission and cross-validation }, author = {Leiner, James and Ramdas, Aaditya}, booktitle = {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics}, pages = {2638--2646}, year = {2024}, editor = {Dasgupta, Sanjoy and Mandt, Stephan and Li, Yingzhen}, volume = {238}, series = {Proceedings of Machine Learning Research}, month = {02--04 May}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v238/leiner24a/leiner24a.pdf}, url = {https://proceedings.mlr.press/v238/leiner24a.html}, abstract = { We introduce a technique called graph fission which takes in a graph which potentially contains only one observation per node (whose distribution lies in a known class) and produces two (or more) independent graphs with the same node/edge set in a way that splits the original graph’s information amongst them in any desired proportion. Our proposal builds on data fission/thinning, a method that uses external randomization to create independent copies of an unstructured dataset. We extend this idea to the graph setting where there may be latent structure between observations. We demonstrate the utility of this framework via two applications: inference after structural trend estimation on graphs and a model selection procedure we term "graph cross-validation"’. } }
Endnote
%0 Conference Paper %T Graph fission and cross-validation %A James Leiner %A Aaditya Ramdas %B Proceedings of The 27th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2024 %E Sanjoy Dasgupta %E Stephan Mandt %E Yingzhen Li %F pmlr-v238-leiner24a %I PMLR %P 2638--2646 %U https://proceedings.mlr.press/v238/leiner24a.html %V 238 %X We introduce a technique called graph fission which takes in a graph which potentially contains only one observation per node (whose distribution lies in a known class) and produces two (or more) independent graphs with the same node/edge set in a way that splits the original graph’s information amongst them in any desired proportion. Our proposal builds on data fission/thinning, a method that uses external randomization to create independent copies of an unstructured dataset. We extend this idea to the graph setting where there may be latent structure between observations. We demonstrate the utility of this framework via two applications: inference after structural trend estimation on graphs and a model selection procedure we term "graph cross-validation"’.
APA
Leiner, J. & Ramdas, A.. (2024). Graph fission and cross-validation . Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 238:2638-2646 Available from https://proceedings.mlr.press/v238/leiner24a.html.

Related Material