Umeň universitet
Faculty of Science and Technology
Department of Computing Science
Umu logo

UMINF reports by Per-┼ke Wedin

1995.01 A New Regularization Method for Rank-deficient Nonlinear Least Squares Problems
1995.02 Regularization Tools for Neural Network Training
1996.03 Regularization Methods for Nonlinear Least Squares Problems. Part I: Exactly Rank-deficient Problems
1996.04 Regularization Methods for Nonlinear Least Squares Problems. Part II: Almost Rank-deficient Problems
1996.05 Regularization Tools for Training Large-Scale Neural Networks
2001.22
2002.16 Interpretation and Practical Use of Error Propagation Matrices
2002.17 First Order Error Analysis of a Linear System of Equations by use of Error Propagation Matrices connected to the Pseudoinverse Solution