Applications Of Regularization To Sem: Shrinking Eigenvalues To Improve Stability Of Covariance Matrices