The classification rules of linear discriminant analysis are defined b
y the true mean vectors and the common covariance matrix of the popula
tions from which the data come. Because these true parameters are gene
rally unknown, they are commonly estimated by the sample mean vector a
nd covariance matrix of the data in a training sample randomly drawn f
rom each population. However, these sample statistics are notoriously
susceptible to contamination by outliers, a problem compounded by the
fact that the outliers may be invisible to conventional diagnostics. H
igh-breakdown estimation is a procedure designed to remove this cause
for concern by producing estimates that are immune to serious distorti
on by a minority of outliers, regardless of their severity. In this ar
ticle we motivate and develop a high-breakdown criterion for linear di
scriminant analysis and give an algorithm for its implementation. The
procedure is intended to supplement rather than replace the usual samp
le-moment methodology of discriminant analysis either by providing ind
ications that the dataset is not seriously affected by outliers (suppo
rting the usual analysis) or by identifying apparently aberrant points
and giving resistant estimators that are not affected by them.