berke@CS.UCLA.EDU (11/29/87)
In many places, Grossberg emphasizes that his concern with real- time issues guided the development of his theories. I greatly admire his work and have read many of his articles. I am in the midst of re-reading several of them, yet I am having difficulty finding anything concrete about real-time behavior. As you know, if you are a Grossberg fan, there is a tremendous amount of material to pour over, all of it interesting. So I have the following questions. If you could answer them for me, I would greatly appreciate it. 1) What is Grossberg's definition of real-time? Instantaeous, combinational, situation-specific, or finitely bounded (similar to linear-bounded automata)? Or, if you have a better definition of real-time, I'd like to hear it and e-talk about it with you. 2) Do Grossberg's systems learn in real-time, or operate in real-time? Whatever they do in real time, do they always do it in real time? Are there any specific paragraphs you can point to that either prove this theoretically or give examples of real- time behavior? I would appreciate any replies, whether they are knowledgeable or naive, specific or tangential, pro- or anti-Grossberg, pro- or anti-real time. Thank you, Peter Berke