Highly Efficient Structural Learning of Sparse Staged Trees

TL;DR

This paper introduces a scalable algorithm for learning sparse staged tree models, enabling efficient structure discovery in complex datasets where previous methods struggled with scalability.

Contribution

The paper presents the first scalable structural learning algorithm for staged trees, focusing on models with limited dependencies to improve efficiency.

Findings

The new algorithm outperforms existing methods in scalability.

Simulation studies demonstrate effective structure learning.

Real-world application confirms practical utility.

Abstract

Several structural learning algorithms for staged tree models, an asymmetric extension of Bayesian networks, have been defined. However, they do not scale efficiently as the number of variables considered increases. Here we introduce the first scalable structural learning algorithm for staged trees, which searches over a space of models where only a small number of dependencies can be imposed. A simulation study as well as a real-world application illustrate our routines and the practical use of such data-learned staged trees.