English

Mixing-Free and Signal-Optimal Learning of Gaussian Graphical Models from Glauber Dynamics

Machine Learning 2026-07-20 v1 Machine Learning Statistics Theory

Abstract

Gaussian graphical model selection is usually studied under independent sampling, but in many applications the data arise as a single trajectory of a dependent stochastic process. We study exact recovery of the graph from one trajectory of random-scan Gaussian Glauber dynamics. Existing techniques for this problem either inherit the mixing time of the chain, which can be super-polynomial in the dimension pp without strong assumptions, or are suboptimal in the minimum normalized edge strength κ\kappa. We propose two algorithms that are mixing-free and attain the κ2\kappa^{-2} dependence of the information-theoretic lower bounds. Both instantiate a shared dueling-neighborhood search meta-algorithm with a local statistic built directly from the update sequence. The first fits a least-squares regression at the updates of each node and recovers the graph from O~(pd2/κ2)\widetilde O(pd^{2}/\kappa^{2}) updates, where dd is the maximum degree. This algorithm's data requirement depends on a local conditioning quantity, but only logarithmically and is provably optimal even when the underlying chain mixes slowly. The second algorithm is based on counting occurences of a specific update pattern and requires O~(pd4/κ2)\widetilde O(pd^{4}/\kappa^{2}) updates, with no dependence on any condition number. The central technical challenge is that both statistics are built from dependent, non-stationary observations. Our analysis tackles this by demonstrating how to extract fresh Gaussian innovations from the update sequence, which yields mixing-free control of appropriate quantities. Neither the algorithms nor their analyses invoke stationarity, a spectral gap, or mixing conditions, and all guarantees hold from an arbitrary initialization.

Keywords

Cite

@article{arxiv.2607.18559,
  title  = {Mixing-Free and Signal-Optimal Learning of Gaussian Graphical Models from Glauber Dynamics},
  author = {Vignesh Tirukkonda and Gautam Dasarathy},
  journal= {arXiv preprint arXiv:2607.18559},
  year   = {2026}
}