American Statistical Association
A measure of explained variation is applied to the Cox proportional hazards model. This measure is useful in understanding the predictive accuracy of the relative risk model. Statistical models have become prevalent in clinical research and the application of these models to aid in the treatment decision process requires confidence in their predictive accuracy. The proposed measure of explained variation is based on the Kullback-Leibler information gain statistic. Its asymptotic and finite sample properties are explored. Applications of the measure to prostate cancer data and as an aid to understanding the qualification of a surrogate marker are illustrated.
|Date:||Wednesday, December 2, 2009|
|Time:||4:00 - 5:00 P.M.|
Memorial Sloan-Kettering Cancer Center
Department of Epidemiology and Biostatistics
307 East 63rd Street
(between First and Second Avenues)
3rd Floor Conference Room
New York, New York
Note: To gain access to the building, please follow the directions by the telephone in the foyer.