RECURSIVE ALGORITHMS FOR PRINCIPAL COMPONENT EXTRACTION

Two new on-line recursive algorithms, namely, the Jacobi recursive pri ncipal component algorithm (JRPCA) and the Gauss-Seidel recursive prin cipal component algorithm (GRPCA), are introduced for the computation of principal components of a slowly varying nonstationary vector stoch astic process. By using these algorithms, the principal components can be adaptively estimated. The speed of convergence of the proposed alg orithms is also discussed. Simulation results show that the proposed a lgorithms have a faster speed of convergence and a better adaptivity w hen compared to other existing methods.