SPECTRAL METHOD FOR SPARSE LINEAR DISCRIMINANT ANALYSIS

摘要

PPROBLEM TO BE SOLVED: To provide a computer implemented method which maximizes candidate solutions to a cardinality-constrained combinatorial optimization problem of sparse linear discriminant analysis. PSOLUTION: A candidate sparse solution vector x with k non-zero elements is input, along with a pair of covariance matrices A, B measuring between-class and within-class covariance of binary input data to be classified and a sparsity parameter k denoting a desired cardinality of a final solution vector. A variational renormalization of the candidate solution vector x is performed with regards to the pair of covariance matrices A, B and the sparsity parameter k to obtain a variance maximized discriminant eigenvector (x) with cardinality k that is locally optimal for the sparsity parameter k and zero-pattern of the candidate sparse solution vector x and is the final solution vector for the sparse linear discriminant analysis optimization problem. PCOPYRIGHT: (C)2008,JPO&INPIT