SEARCH AND INFORMATION THEORY. THE LOGICAL BASIS OF COMBAT SIMULATION
Abstract
When a probability distribution of a target is given, the problem of detecting it by a given method involving the progressive expenditure of effort, will be harder or easier according to the nature of the distribution. A natural measure of the degree of difficulty of detection by the method in question is the expected value of the quantity of effort used up to the (random) moment of detection--assuming that the most effective use of the effort is made at each moment. The first object of this investigation is to show the connection between the present concept when the search is done as a procedure of successive dichotomies until localization--to a pre-stated order of precision--is obtained. The connection with the classical theory of information (entropy and cross- entropy) is shown. The second object of this investigation is to apply the method to the case in which the law of random search is applied progressively and optimally to the search for a target in a linear, planar, or spacial region (or in r-space).
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1968
- Accession Number
- AD0687534
Entities
People
- Bernard O. Koopman
Organizations
- Columbia University