| Document Date: 2008-07-16 21:59:51
Open Document File Size: 186,48 KBShare Result on Facebook
City Natick / Hsinchu / Washington / D.C. / Como / Derivative / / Company Princeton University Press / MathWorks Inc. / Neural Information Processing Systems / BP / Neural Networks / / Country Taiwan / Italy / / / Facility Prentice Hall / National Tsing Hua University / / IndustryTerm ve algorithms / adaptive networks / direct dogleg algorithms / representative batch-mode algorithms / ective algorithms / derivative-computing process / gradient algorithms / online-mode steepest descent / deterministic algorithm / batch-mode direct dogleg trust-region algorithm / direct dogleg algorithm / dogleg algorithms / learning algorithms / est descent-type algorithm / / Organization World Congress / Princeton University / National Tsing Hua University / Univ. of California / / Person James W. Demmel / S. Roger Jang / Stuart E. Dreyfus / Eiji Mizutani / Shing Roger Jang / H. Kelley Adjoint / / Position representative / / ProgrammingLanguage J / / ProvinceOrState South Dakota / California / Massachusetts / / Technology ve algorithms / est descent-type algorithm / direct dogleg algorithm / existing NNlearning algorithms / Neural Network / MLP-learning algorithms / deterministic algorithm / gradient algorithms / direct dogleg algorithms / algorithm Dogleg algorithm / dogleg algorithms / quasi-Newton algorithms / two representative batch-mode algorithms / NN algorithms / Five algorithms / batch-mode direct dogleg trust-region algorithm / /
SocialTag |
On dynamic programming-like recursive gradient formula for alleviating MLP hidden-node saturation in the parity problem Eiji Mizutani 1,