Australasian Plant Pathology

, Volume 43, Issue 3, pp 267–286

The structure of diagnostic information

Article

DOI: 10.1007/s13313-013-0267-2

Cite this article as:
Hughes, G. & McRoberts, N. Australasian Plant Pathol. (2014) 43: 267. doi:10.1007/s13313-013-0267-2

Abstract

Diagnosis is characterized as an exercise in classification, where the task is to assign a crop to a risk group as a basis for evidence-based crop protection decision making. Underlying the process of diagnostic decision making is Bayesian updating of probabilities. Alongside updating of probabilities, assessments of diagnostic information allow further description of the characteristics of diagnostic tests, and of the predictions made on the basis of test outcomes. This is illustrated analytically, graphically (by means of iso-information contour plots and information graphs) and by discussion of example epidemiological scenarios.

Keywords

Bayes’ theorem Entropy Expected mutual information Specific information Relative entropy Weight of evidence 

Copyright information

© Australasian Plant Pathology Society Inc. 2014

Authors and Affiliations

  1. 1.Crop and Soil SystemsSRUCEdinburghUK
  2. 2.Plant Pathology DepartmentUniversity of CaliforniaDavisUSA

Personalised recommendations