Mathematics, Department of
Document Type
Article
Date of this Version
7-22-2021
Citation
(2023) SIAM Journal on Computing, 52 (3), pp. 761-793. DOI: 10.1137/22m1489678
Abstract
We provide finite sample guarantees for the classical Chow-Liu algorithm (IEEE Trans. Inform. Theory, 1968) to learn a tree-structured graphical model of a distribution. For a distribution P on Σn and a tree T on n nodes, we say T is an ε-approximate tree for P if there is a T-structured distribution Q such that D(P || Q) is at most ε more than the best possible tree-structured distribution for P. We show that if P itself is tree-structured, then the Chow-Liu algorithm with the plug-in estimator for mutual information with eO (|Σ|3nε−1) i.i.d. samples outputs an ε-approximate tree for P with constant probability. In contrast, for a general P (which may not be tree-structured), Ω(n2ε−2) samples are necessary to find an ε-approximate tree. Our upper bound is based on a new conditional independence tester that addresses an open problem posed by Canonne, Diakonikolas, Kane, and Stewart (STOC, 2018): we prove that for three random variables X, Y, Z each over Σ, testing if I(X; Y | Z) is 0 or ≥ ε is possible with Õ(|Σ|3/ε) samples. Finally, we show that for a specific tree T , with Õ(|Σ|2nε−1) samples from a distribution P over Σn, one can efficiently learn the closest T -structured distribution in KL divergence by applying the add-1 estimator at each node.
Comments
Used by permission.