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.

Let W be a Wiener process. For r0 and T0 let IW(T,r)2 denote the minimal value of the energy ∫0Th′(t)2dt taken among all absolutely continuous functions h(⋅) defined on [0,T], starting at zero and satisfying W(t)−r≤h(t)≤W(t)+r,0≤t≤T. The function minimizing energy is a taut...