Scalable Bayesian Network Structure Learning Using Tsetlin Machine to Constrain the Search Space

TL;DR

This paper introduces a scalable Bayesian network structure learning method that uses the Tsetlin Machine to efficiently constrain the search space, significantly reducing computational time while maintaining accuracy.

Contribution

The study presents a novel TM-based approach that leverages significant literals to perform CI tests, improving efficiency over traditional PC algorithms in large-scale datasets.

Findings

Reduces computational complexity significantly.

Maintains competitive accuracy in causal discovery.

Outperforms traditional methods in large datasets.

Abstract

The PC algorithm is a widely used method in causal inference for learning the structure of Bayesian networks. Despite its popularity, the PC algorithm suffers from significant time complexity, particularly as the size of the dataset increases, which limits its applicability in large-scale real-world problems. In this study, we propose a novel approach that utilises the Tsetlin Machine (TM) to construct Bayesian structures more efficiently. Our method leverages the most significant literals extracted from the TM and performs conditional independence (CI) tests on these selected literals instead of the full set of variables, resulting in a considerable reduction in computational time. We implemented our approach and compared it with various state-of-the-art methods. Our evaluation includes categorical datasets from the bnlearn repository, such as Munin1, Hepar2. The findings indicate that the proposed TM-based method not only reduces computational complexity but also maintains competitive accuracy in causal discovery, making it a viable alternative to traditional PC algorithm implementations by offering improved efficiency without compromising performance.