Zero-suppressed binary decision diagrams (ZDDs) are a data structure representing Boolean functions, and one of the most successful variants of binary decision diagrams (BDDs). On the other hand, BDDs are also called branching programs in computational complexity theory, and have been studied as a computation model. In this paper, we consider ZDDs from the viewpoint of computational complexity theory. Firstly, we define zero-suppressed branching programs, which actually have the same definition to (unordered) ZDDs, and consider the computational power of zero-suppressed branching programs. Secondly, we attempt to generalize the concept of zero-suppression. We call the basic idea of ZDDs zero-suppression. We show that zero-suppression can be applied to other two classical computation models, decision trees and Boolean formulas.
展开▼
