Neural network based approach for eigenstructure extraction

A new neural network approach for eigenstructure extraction is proposed. It is based on the constraint optimal problem. It can be used to estimate the largest eigenvector of the covariance matrix adaptively and efficiently. In order to extract the other eigenvectors, a covariance matrix series is constructed. A cost function is discussed for extracting the largest eigenvectors of a covariance matrix, and then, considering the cost function as a bridge, a high-order neural network is introduced to extract the largest eigenvector. Theoretical analysis and simulations show that the proposed approach is efficient and direct for eigenstructure extraction