Springer, Lecture Notes in Computer Science, p. 218-231, 2005
DOI: 10.1007/11556114_14
Full text: Download
Information theory has long provided a mathematical framework for statistically evaluating the diversity a given signal is able to produce. This paper applies information theory to multi-dimensional spaces, discusses the limits of conventional measures of diversity, and introduces a spatial component into the measure of diversity. This leads to a spatial form of diversity that provides a novel way to study spatial structures where categories and distances are amongst key factors in the analysis. The approach is exemplified with spatial configuration samples.