American Statistical Association
Calibration refers to the agreement between model-predicted probabilities and observed frequencies. It is usually evaluated by grouping the predicted probabilities and computing the observed proportion for these groups. If predicted probabilities and observed proportions are on the 45-degree line then the model is properly calibrated. In this talk, I will present local regression as an alternative to grouping the predicted probabilities and discuss the use of the area between the local regression estimate and 45-degree line as an index of calibration. Examples will focus on logistic and Cox regression.
|Date:||Wednesday, May 30, 2007|
|Time:||4:00 P.M. - 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.