【作者】 孔敏； 朱大奇； 张伟；

【Author】 Kong Min Zhu Daqi Zhang wei(School of Communication and control Engineering, Southern Yangtze University, Wu Xi, China, 214036)

【机构】 江南大学通信与控制工程学院；

【摘要】 为提高CMAC(cerebellar model articulation controllers)神经网络在线学习的快速性和准确性,提出了一种平衡学习的概念,设计了一种改进的CMAC学习算法。在常规的CMAC中,误差的校正值被平均地分配给所有激活存储单元,而不管这些存储单元的可信度。在改进的CMAC中,利用激活单元先前学习次数作为可信度。其误差校正值与激活单元前学习次数的负k次方成比例。仿真结果表明当k为一适当数值时,改进CMAC有较快的学习速度和较高的精度,尤其是在神经网络的初始学习阶段。更多还原

【Abstract】 In order to improve online learning speed and accuracy of CMAC , An improved CMAC neural network model based on balanced learning concept are designed. In the conventional CMAC learning scheme, the correcting amounts of errors are equally distributed into all addressed hypercubes, regardless the credibility of those hypercubes. The proposed improved learning approach is to use the learned times of the addressed hypercubes as the credibility (confidence) of learned values.With this idea, the learning speed can indeed become very fast when k is a best value in the early learning stage especially.更多还原