Bayesian learning of forest and tree graphical models (2108.13992v1)

Abstract: In Bayesian learning of Gaussian graphical model structure, it is common to restrict attention to certain classes of graphs and approximate the posterior distribution by repeatedly moving from one graph to another, using MCMC or methods such as stochastic shotgun search (SSS). I give two corrected versions of an algorithm for non-decomposable graphs and discuss random graph distributions, in particular as prior distributions. The main topic of the thesis is Bayesian structure-learning with forests or trees. Restricting attention to these graphs can be justified using theorems on random graphs. I describe how to use the Chow$\unicode{x2013}$Liu algorithm and the Matrix Tree Theorem to find the MAP forest and certain quantities in the posterior distribution on trees. I give adapted versions of MCMC and SSS for approximating the posterior distribution for forests and trees, and systems for storing these graphs so that it is easy to choose moves to neighbouring graphs. Experiments show that SSS with trees does well when the true graph is a tree or sparse graph. SSS with trees or forests does better than SSS with decomposable graphs in certain cases. Graph priors improve detection of hubs but need large ranges of probabilities. MCMC on forests fails to mix well and MCMC on trees is slower than SSS. (For a longer abstract see the thesis.)