Receiver Operator Characteristic Curves in the evaluation of medical diagnosis
 
 
Description:  The origin of Receiver Operating Characteristic (ROC) analysis can be traced to the theory of statistical decision making, related to applications in signal detection and psychology and, over the years, has become a powerful tool for measuring diagnostic tests performance in medicine. A diagnostic test aims to classify individuals into two classes: normal and abnormal. Since classification may be imperfect, there exists the possibility of making one of two types of errors: to classify an normal individual as abnormal and vice versa. For a given test, the compromise between the False Positive Fraction (FPF) and the True Positive Fraction (TPF) can be graphically presented through a ROC curve. Several examples of the application of ROC curves in medical diagnosis will be presented.
Date:  2012-06-06
Start Time:   14:30
Speaker:  Pedro Oliveira (Instituto de Ciências Biomédicas Abel Salazar, Univ. Porto)
Institution:  Instituto de Ciências Biomédicas Abel Salazar, Univ. Porto
Place:  Sala 5.5
Research Groups: -Analysis
See more:   <Main>  
 
© Centre for Mathematics, University of Coimbra, funded by
Science and Technology Foundation
Powered by: rdOnWeb v1.4 | technical support