Abstract
Diagrammatic formats are useful for summarizing the processes of evaluation and comparison of forecasts in plant pathology and other disciplines where decisions about interventions for the purpose of disease management are often based on a proxy risk variable. We describe a new diagrammatic format for disease forecasts with two categories of actual status and two categories of forecast. The format displays relative entropies, functions of the predictive values that characterize expected information provided by disease forecasts. The new format arises from a consideration of earlier formats with underlying information properties that were previously unexploited. The new diagrammatic format requires no additional data for calculation beyond those used for the calculation of a receiver operating characteristic (ROC) curve. While an ROC curve characterizes a forecast in terms of sensitivity and specificity, the new format described here characterizes a forecast in terms of relative entropies based on predictive values. Thus it is complementary to ROC methodology in its application to the evaluation and comparison of forecasts.
Highlights
Forecasting using two categories of actual status and two categories of forecast is common in many scientific and technical applications where evidence-based risk assessment is required as a basis for decision-making, including plant pathology and clinical medicine
We wish to highlight that performance metrics conditioned on forecast outcomes have a useful role in the overall evaluation of diagnostic tests and disease forecasters; second, bearing in mind the first aim, we wish to demonstrate that performance metrics based on information theoretic quantities can help distinguish characteristics of such tests and forecasters that may not be apparent from probability-scale metrics
Calculating the new diagrammatic format that we describe here has the same data requirements as the calculation of the receiver operating characteristic (ROC) curve, but relates to relative entropy, an information theoretic metric that quantifies the expected amount of diagnostic information consequent on probability revision from prior to posterior arising from application of a disease forecaster
Summary
Forecasting using two categories of actual status and two categories of forecast is common in many scientific and technical applications where evidence-based risk assessment is required as a basis for decision-making, including plant pathology and clinical medicine. The statistical evaluation of probabilistic disease forecasts often involves the calculation of metrics defined conditionally on actual disease status. For the purpose of disease management decision making, metrics defined conditionally on forecast outcomes (i.e., predictive values) are of interest, these are less frequently reported. We introduce a new diagrammatic format for disease forecasts with two categories of actual status and two categories of forecast. The new diagrammatic format we introduce is intended to provide a generic approach that can applied in any suitable context
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have