NEW APPLICATIONS OF FAILURE FUNCTIONS.

Abstract

Presented are several algorithms whose operations are governed by a principle of failure functions. When searching for an extremal value within a sequence, it suffices to consider only the subsequence of items each of which is the first possible improvement of its predecessor. These algorithms are more efficient than their more traditional counterparts.