FURTHER INVESTIGATION OF THE EFFECTS OF REDUCED INPUT DATA FIDELITY UPON THE DETERMINATION OF POSTERIOR PROBABILITIES IN A SIMULATED THREAT-DIAGNOSIS SYSTEM.
Abstract
This is the fifth in a series of experiments on Bayesian diagnostic systems. In this experiment two procedures for obtaining a posteriori (P (H/D)) probability estimates were compared under conditions of low fidelity of input data and various levels of time-stress. In one procedure a computer aggregated the subjects' a priori (P (D/H)) estimates. In the alternate procedure the a posteriori probabilities were estimated by the subjects without computer aid. The results favor the computer-aided procedure and tend to support the use of automated Bayesian hypothesis-selection procedures in diagnostic systems. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1965
- Accession Number
- AD0631781
Entities
People
- David A. Schum
- Irwin L. Goldstein
- Jack F. Southard
Organizations
- Ohio State University