Improved Upper Bounds for Determinizing NIDPDAs with Limited Nondeterminism
摘要
Determinizing a nondeterministic input-driven pushdown automaton (NIDPDA) of size n results in the worst case in a machine of size \(2^{\varTheta (n^2)}\) (R. Alur, P. Madhusudan, “Adding nesting structure to words”, J.ACM 56(3), 2009). A significantly improved determinization upper bound is known for k-path NIDPDAs that can have at most k computations on any input. We improve the known upper bound of determinization of k-path NIDPDAs by, roughly speaking, making the deterministic input-driven machine (DIDPDA) remember the shape of the tree in which the nondeterministic choices were made. Also we provide upper bounds for converting k-path universal input driven pushdown automata to DIDPDAs and establish lower bounds.