Explanation-Based Learning in Logic Circuit Design

Abstract

Our research is focused on a technique known as explanation-based learning. We are studying this technique by applying it to the task of learning to design logic circuits. This approach involves utilizing domain knowledge to analyze example circuits and produce generalized circuit designs. In particular, proofs of design correctness are used to guide the process of generalization. This method has been demonstrated in a program which generalizes a circular shift register into a schema describing devices capable of computing arbitrary permutations.