Information ambiguity
Samar Agnihotri, H. S. Jamadagni · 2008
We propose a quantity called information ambiguity that plays the same role in the worst-case information-theoretic analyses as the well-known notion of information entropy performs in the corresponding average-case analyses. We prove various properties of information ambiguity and illustrate its usefulness in performing the worst-case analysis of a variant of distributed source coding problem.