ABSTRACT
This paper describes briefly the theory and character of singular value decomposition of matrix emphasizes the idea of solving the over-determined 、ill-conditioned equation series, proposes two SVD based algorithms and gives a comparison of the two projects. after discusses the signals what the equations embody, samples from the signals to form the matrix by the data, use the MATLAB to analyzed the outputs of these algorithms for processing the observed data. And select the project which is better. selects the singular value to optimize the figure of the signals while using the singular value decomposition to analyze the data, finally, under the same conditions, I analyzed the data using the least-squares solution ,makes the conclusion that singular value decomposition is more steady than the least-squares solution, so the singular value decomposition method is recommended in data processing.
Keywords:singular value decomposition(SVD); ill-conditioned equation series;
least-squares solution ; stability