Two entropies of a generalized sorting problem

作者:

Highlights:

摘要

In this paper we introduce a class of generalized sorting (ordering) problems called “classifications.” To each “classification,” we associate two quantities: informational entropy (average information quantity) and operational entropy (measure of computational complexity, that is, number of comparisons necessary to “classify” a given sequence of items). The relationship between these quantities is discussed. For a certain classification involving n items, its operational entropy is shown to be approximately n·log2n although its informational entropy is constantly equal to 1, independent of the number of items n.

论文关键词:

论文评审过程:Received 19 September 1972, Available online 31 December 2007.

论文官网地址:https://doi.org/10.1016/S0022-0000(73)80038-8