Uniform characterizations of non-uniform complexity measures

Abstract

Non-uniform complexity measures originated in automata and formal languages theory are characterized in terms of well-known uniform complexity classes. The initial index of languages is introduced by means of several computational models. It is shown to be closely related to context-free cost, boolean circuits, straight line programs, and Turing machines with sparse oracles and time or space bounds.