Author:
Großmann Gerrit,Zimmerlin Julian,Backenköhler Michael,Wolf Verena
Abstract
Abstract
Problem setting
Stochastic dynamical systems in which local interactions give rise to complex emerging phenomena are ubiquitous in nature and society. This work explores the problem of inferring the unknown interaction structure (represented as a graph) of such a system from measurements of its constituent agents or individual components (represented as nodes). We consider a setting where the underlying dynamical model is unknown and where different measurements (i.e., snapshots) may be independent (e.g., may stem from different experiments).
Method
Our method is based on the observation that the temporal stochastic evolution manifests itself in local patterns. We show that we can exploit these patterns to infer the underlying graph by formulating a masked reconstruction task. Therefore, we propose (raph nference etwork rchitecture), a machine learning approach to simultaneously learn the latent interaction graph and, conditioned on the interaction graph, the prediction of the (masked) state of a node based only on adjacent vertices. Our method is based on the hypothesis that the ground truth interaction graph—among all other potential graphs—allows us to predict the state of a node, given the states of its neighbors, with the highest accuracy.
Results
We test this hypothesis and demonstrate ’s effectiveness on a wide range of interaction graphs and dynamical processes. We find that our paradigm allows to reconstruct the ground truth interaction graph in many cases and that outperforms statistical and machine learning baseline on independent snapshots as well as on time series data.
Publisher
Springer Science and Business Media LLC
Subject
Computational Mathematics,Computer Networks and Communications,Multidisciplinary
Reference56 articles.
1. Amini H, Cont R, Minca A (2016) Resilience to contagion in financial networks. Math Financ 26(2):329–365
2. Bak P, Chen K, Tang C (1990) A forest-fire model and some thoughts on turbulence. Phys Lett A 147(5–6):297–300
3. Barman S, Kwon Y-K (2018) A boolean network inference from time-series gene expression data using a genetic algorithm. Bioinformatics 34(17):927–933
4. Bengio E, Jain M, Korablyov M, Precup D, Bengio Y(2021) Flow network based generative models for non-iterative diverse candidate generation. Adv Neural Inf Process Syst 34
5. Campbell A, Gurin G, Miller WE (1945) The voter decides