A LEARNING-METHOD FOR NEURAL NETWORKS BASED ON A PSEUDOINVERSE TECHNIQUE

A theoretical formulation of a fast learning method based on a pseudoi nverse technique is presented. The efficiency and robustness of the me thod are verified with the help of an Exclusive OR problem and a dynam ic system identification of a linear single degree of freedom mass-spr ing problem. It is observed that, compared with the conventional backp ropagation method, the proposed method has a better convergency rate a nd a higher degree of learning accuracy with a lower equivalent learni ng coefficient. It is also found that unlike the steepest descent meth od, the learning capability of which is dependent on the value of the learning coefficient eta, the proposed pseudoinverse based backpropaga tion algorithm is comparatively robust with respect to its equivalent variable learning coefficient. A combination of the pseudoinverse meth od and the steepest descent method is proposed for a faster, more accu rate learning capability. (C) 1996 John Wiley & Sons, Inc.