How to calculate correlation measure for logistic regression?

by tatami   Last Updated July 17, 2017 10:19 AM

Zheng (2000) recommended the use of correlation measures to summarize predictive power. I would like to use this measure to compare nested descriptive regression models. He mentioned a few types of correlation measure, mainly sample estimator; a jack-knife estimator; a modified jack-knife estimator, and a leave-one-out cross-validation estimator. How do I calculate them and their bootstrap confidence intervals in R?

Related Questions