Learning Minimum-Size BDDs: Towards Efficient Exact Algorithms

Abstract

Binary decision diagrams (BDDs) are widely applied tools to compactly represent labeled data as directed acyclic graphs; for efficiency and interpretability reasons small BDDs are preferred. Given labeled data, minimizing BDDs is NP-complete and thus recent research focused on the influence of parameters such as the solution size $s$ on the complexity [Ordyniak et al., AAAI 2024]. Our main positive result is an algorithm that is efficient if in particular $s$, the domain size $D$, and the Hamming distance between any two data points is small, improving on previous running-time bounds. This algorithm is inspired by the witness-tree paradigm that was recently successful for computing decision trees [Komusiewicz et al., ICML 2023], whose extension to BDDs was open. We extend our algorithmic results to the case where we allow a small number of misclassified data points and complement them with lower bounds that show that the running times are tight from multiple points of view. We show that our main algorithm holds practical promise by providing a proof-of-concept implementation.