基于ART的RBF网络结构设计.pdfVIP

绗?9 鍗 绗?0 鏈 鎺 鍒 涓 鍐 绛 2014 骞 10 鏈 Vol. 29 No. 10 Control and Decision Oct. 2014 鏂囩珷缂栧彿: 1001-0920 (2014) 10-1876-05 DOI: 10.13195/j.kzyjc.2013.0945 鍩轰簬ART 鐨凴BF 缃戠粶缁撴瀯璁捐 钂 瑗? 涔斾繆椋? 闊╃孩妗 (鍖椾含宸ヤ笟澶у鐢靛瓙淇℃伅涓庢帶鍒跺伐绋嬪闄紝鍖椾含100124) 鎽 瑕? 閽堝寰勫悜鍩哄嚱鏁?RBF) 缃戠粶闅愬眰缁撴瀯闅句互纭畾鐨勯棶棰? 鍩轰簬鑷€傚簲鍏辨尟鐞嗚(ART) 缃戠粶鑹ソ鐨勫湪绾垮垎绫荤壒 鎬? 鎻愬嚭涓€绉峈BF 缃戠粶缁撴瀯璁捐绠楁硶. 璇ョ畻娉曞皢ART 缃戠粶鐨勮仛绫荤壒鎬х敤浜嶳BF 缃戠粶缁撴瀯璁捐涓? 閫氳繃瀵硅緭鍏ュ悜閲忎笌 宸插瓨妯″紡鐨勭浉浼煎害姣旇緝灏嗚緭鍏ュ悜閲忚繘琛屽垎绫? 纭畾闅愬惈灞傝妭鐐逛釜鏁板拰鍒濆鍙傛暟, 浣跨綉缁滃叿鏈夌簿绠€鐨勭粨鏋? 瀵瑰吀鍨嬮潪绾 鎬у嚱鏁伴€艰繎鐨勪豢鐪熺粨鏋滆〃鏄? 鎵€鎻愬嚭鐨勭粨鏋勫叿鏈夊揩閫熺殑瀛︿範鑳藉姏鍜岃壇濂界殑閫艰繎鑳藉姏. 鍏抽敭璇岤 鑷€傚簲鍏辨尟鐞嗚缃戠粶锛涘緞鍚戝熀鍑芥暟绁炵粡缃戠粶锛涚粨鏋勮璁 涓浘鍒嗙被鍙耳 TP273 鏂囩尞鏍囧織鐮? A RBF Neural network based on ART neural network MENG Xi, QIAO Jun-fei, HAN Hong-gui (College of Electronic Information and Control Engineering锛孊eijing University of Technology 锛孊eijing 100124, China 锛嶤orrespondent 锛歈IAO Jun-fei锛孍-mail 锛歫unfeiq@bjut.edu.cn) Abstract: For the problem that it is dif铿乧ult to determine the hidden layer structure of the radial basis function(RBF) neural network, based on the good online classi铿乪d characteristic of adaptive resonance theory(ART) neural network, a self-organizing RBF neural network structure design algorithm is proposed. The algorithm uses the clustering characteristic of ART neural network to design the RBF neural network structure. Through the similarity comparison of input vector, the number of the hidden layer nodes and initial parameters are determined, so that the network has simpli铿乪d structure. The experiment results show that the proposed structure has a smaller number of nodes, fast learning speed and better approximation abi