人工脑模式识别种新的神经网络算法.pdf

j¨¨¨¨¨， ㈣8帅2帅7 Abstract 眦Y 帆2洲3呲3㈣8 At ofthevariousindustrieson present，therequirements the patternrecognition aremoreand andmore more，more ofthetraditional technology higher．However,some are the pattern absenceof ofthe recognitiontechnologiesexperiencing priorknowledge use of“theconnectedcharacterof correct homologue”，the rateis difficultto recognitionvery the bottleneck． improve This is on basedtheNationalNaturalScienceFoundation paper on‘‘artificial project braininformationnew neuralnetworkmodel”(No，“artificial processing brain newneuralnetwork patternrecognition modelbasedon connectedness” homologue (NO．61 viewofthe 272077)，in massmorethanclass patternrecognitionproblems， the“theconnectedcharacterof foraclass through of homologue”ofcognition，proposed artificialbrain onblock orderedneural pattemrecognition parallelpriority network in block orderedRBFneuralnetwork algorithm，rais