基于熵误差函数Pi-Sigma神经网络在线梯度学习算法  

Online Gradient Learning Algorithm Based on Pi-Sigma Neural Network with Entropy Error Function

作　　者：郑凤姣 董青梅 屈荣国 范钦伟 ZHENG Fengjiao;DONG Qingmei;QU Rongguo;FAN Qinwei(Xi'an Polytechnic University,Xi'an Shaanxi 710600,China)

机构地区：[1]西安工程大学,陕西西安710600

出　　处：《信息与电脑》2023年第6期120-122,共3页Information & Computer

摘　　要：Pi-Sigma神经网络因具有较强的非线性处理能力和较高的学习效率而受到广泛的关注。目前,该网络已经应用于函数逼近、信号处理、模式识别等领域。因此,提出了一种基于熵误差函数的Pi-Sigma神经网络在线学习算法,通过函数逼近及多维数据集分类实验验证了该算法的有效性。结果表明,所提算法在函数逼近、训练精度、测试精度和效率方面均有较好的表现。Pi-Sigma neural network has been widely concerned because of its strong nonlinear processing ability and high learning efficiency.The network has now been applied to function approximation,signal processing,pattern recognition and many other areas.In this paper,a Pi-Sigma neural network online learning algorithm based on entropy error function is proposed.The effectiveness of the algorithm is verified through function approximation and multidimensional data set classification experiments.The results show that the proposed algorithm has good performance in function approximation,training accuracy,testing accuracy and efficiency.

关 键 词：PI-SIGMA神经网络 在线梯度算法 熵误差函数 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]