Hamilton.ES@XEROX.ARPA (04/24/84)
From: Bruce Hamilton <Hamilton.ES@XEROX.ARPA> The research described below sounds closer to what I had in mind when I raised this issue a couple of weeks ago, as opposed to the automata-theoretic responses I tended to get. --Bruce [For more leads on learning "systems containing state variables", readers should look into that branch of control theory known as system identification. Be prepared to deal with some hairy mathematical notation. -- KIL] Date: 24 Apr 84 11:39 PST From: mittal.pa Subject: Reminder: CSDG Today The CSDG today will be given by Tom Dietterich, Stanford University, based on his thesis research work. Time etc: Tuesday, Apr. 24, 4pm, Twin Conf. Rm (1500) Learning About Systems That Contain State Variables It is difficult to learn about systems that contain state variables when those variables are not directly observable. This talk will present an analysis of this learning problem and describe a method, called the ITERATIVE EXTENSION METHOD, for solving it. In the iterative extension method, the learner gradually constructs a partial theory of the state-containing system. At each stage, the learner applies this partial theory to interpret the I/O behavior of the system and obtain additional constraints on the structure and values of its state variables. These constraints trigger heuristics that hypothesize additional internal state variables. The improved theory can then be applied to interpret more complex I/O behavior. This process continues until a theory of the entire system is obtained. Several conditions sufficient to guarantee the success of the method will be presented. The method is being implemented and applied to the problem of learning UNIX file system commands by observing a tutorial interaction with UNIX.