摘要
Lee的复域多值双向联想记忆模型(complex domain bidirectional associative memory,简称CDBAM)不仅将Kosko的实域BAM(bidirectional associative memory)推广至复域,而且推广至多值情形,以利于多值模式(如灰级图像等)间的联想.在此基础上,提出了一个新的推广模型:复域内连式多值双向联想记忆模型(intraconnected CDBAM,简称ICDBAM),通过定义的能量函数证明了它在同步与异步更新方式下的稳定性,从而保证所有训练样本对成为其稳定点,克服了CDBAM所存在的补码问题.计算机模拟证明了该模型比CDBAM具有更高的存储容量和更好的纠错性能.
Lees multivalued bidirectional associative memory operating on a complex domain (CDBAM) extends Koskos BAM (bidirectional associative memory) not only to complex domain but also to multivalued situation in favor of associating on multivalued model. Based on the CDBAM, a new extended model, ICDBAM (intraconnected multivalued bidirectional associative memory operating on a complex domain), is presented in this paper. The stability of the new model in synchronous and asynchronous updating modes, is proven by defining an energy function such that it can ensure all the training pattern pairs to become its asymptotically stable points. In addition, it eliminates CDBAM抯 complementing encoding problem. The computer simulations show that the proposed model has higher storage capacity and better error-correcting capability than CDBAM.
出处
《软件学报》
EI
CSCD
北大核心
2002年第3期433-437,共5页
Journal of Software
基金
国家自然科学基金资助项目(69701004)
国家教育部青年骨干教师资助项目
南京大学计算机软件新技术国家重点实验室基金资助项目~~
关键词
神经网络
能量函数
复合域
人工智能
内连式复值双向联想记忆模型
性能分析
bidirectional associative memory
neural networks
energy function
complex domain
multivalued associative memory
intraconnection
