Comparison-based search in the presence of errors

Abstract

The classic problem of searching an ordered list becomes more complicated when the answers to queries are not reliable. In the prefix-bounded error model, the number of incorrect responses at no point exceeds r times the number of questions, where r is a fixed positive fraction strictly less than ~. We present efficient algorithms for searching in this error model, and we also establish nontrivial lower bounds for sorting, maxima finding, and related problems.