Optimal discrimination linear mapping (Fisher mapping, LDA)
W = FISHERM(A,N,ALF)
Finds a mapping of the labeled dataset A onto an N-dimensional linear subspace such that it maximises the the between scatter over the within scatter (also called the Fisher mapping [1 or LDA]). Note that N should be less than the number of classes in A. If supplied, ALF determines the preserved variance in the prewhitening step (i.e. removal of insignificant eigenvectors in the within-scatter, the EFLD procedure ), see KLMS.
The resulting mapping is not orthogonal. It may be orthogonalised by ORTH.
 K. Fukunaga, Introduction to statistical pattern recognition, 2nd ed., Academic Press, New York, 1990.  C. Liu and H. Wechsler, Robust Coding Schemes for Indexing and Retrieval from Large Face Databases, IEEE Transactions on Image Processing, vol. 9, no. 1, 2000, 132-136.