BALANCED BOOTSTRAP ESTIMATORS FOR THE PROBABILITY OF MISCLASSIFICATIONS IN DISCRIMINANT ANALYSIS

  • I W. MANGKU Bogor Agricultural University

Abstract

In this paper we propose new error rate estimators based on balanced bootstrap technique, which are expected to per- form better than the existing estimators. These estimators can be computed by means of separate or mixture resampling methodol- ogy. We consider both of them.

Downloads

Download data is not yet available.
Published
2007-07-01