02659nam 2200601 a 450 991095723030332120251116181846.01-62257-103-7(CKB)2550000001043311(EBL)3021387(SSID)ssj0000880417(PQKBManifestationID)12428830(PQKBTitleCode)TC0000880417(PQKBWorkID)10895997(PQKB)10699939(MiAaPQ)EBC3021387(Au-PeEL)EBL3021387(CaPaEBR)ebr10683045(OCoLC)839886484(BIP)32841682(EXLCZ)99255000000104331120110225d2011 uy 0engur|n|---|||||txtccrArtificial neural network training and software implementation techniques /Ali Kattan, Rosni Abdullah and Zong Woo Geem1st ed.Hauppauge, N.Y. Nova Science Publishersc20111 online resource (68 p.)Computer networksNovinkaDescription based upon print version of record.1-61122-990-1 Includes bibliographical references (p.[43]-53) and index.Feed-forward neural networks -- FFANN software simulation -- FFANN training concept -- Trajectory-driven training paradigm -- Evolutionary-based training paradigm -- FFANN simulation utilizing graphic-processing units.Artificial neural networks (ANN) are widely used in diverse fields of science and industry. Though there have been numerous techniques used for their implementations, the choice of a specific implementation is subjected to different factors including cost, accuracy, processing speed and overall performance. Featured with synaptic plasticity, the process of training is concerned with adjusting the individual weights between each of the individual ANN neurons until we can achieve close to the desired output. This book introduces the common trajectory-driven and evolutionary-based ANN training algorithms.Computer networks series.Novinka.Neural networks (Computer science)Neural networks (Computer science)006.3/2Kattan Ali1861271Abdullah Rosni1861272Geem Zong Woo1133777MiAaPQMiAaPQMiAaPQBOOK9910957230303321Artificial neural network training and software implementation techniques4467362UNINA