
 Finite Automata 
  In a stochastic finite automaton (SFA), f(q,a) and h(q) are random functions and Pr(f(q,a)q,a) and Pr(h(q)) define the state transition and output probabilities respectively. 
  In a probabilistic finite automata (PFA), the output in state q is a random function of q, and the state following q is a deterministic function of q and the output in q. 
  PFSA is a PFA in which each state in Q has a unique label corresponding to a sequence (string) of outputs and, moreover, the label for f(q,h(q)) is a suffix of the sequence defined by concatenating the output h(q) to the label for q. 
 www.cs.brown.edu /research/ai/dynamics/tutorial/Documents/FiniteAutomata.html (1520 words) 
