Fisher’s linear discriminant function might be difficult to estimate, when data come from a semiparametric, finite mixture model. We propose an estimator based on the singular value decomposition of the third standardized cumulant. The estimator is consistent when sampling from a mixture of...

When data come from a mixture of two weakly symmetric distributions with different means and proportional covariances, we estimate the best linear discriminant projection with the projection of the standardized data onto the direction of a vector-valued measure of skewness.