A critical aspect in the problem of inductive inference is the number
of examples needed to accurately infer a Boolean function from positiv
e and negative examples. In this paper, we develop an approach for der
iving a sequence of examples for this problem. Some computer experimen
ts indicate that, on the average, examples derived according to the pr
oposed approach lead to the inference of the correct function consider
ably faster than when examples are derived in a random order.