基于改进MRII的Madaline网络学习算法

doi:10.3969/j.issn.1006-2475.2016.02.011

计算机与现代化

基于改进MRII的Madaline网络学习算法

河海大学计算机与信息学院智能科学与技术研究所，江苏南京211100

收稿日期:2015-09-21 出版日期:2016-03-02 发布日期:2016-03-03
作者简介: 郑星（1990-）,男，江苏镇江人，河海大学计算机与信息学院智能科学与技术研究所硕士研究生，研究方向：机器学习，神经网络。
基金资助:
国家自然科学基金资助项目（60971088）

A Madaline Network Learning Algorithm Based on Improved MRII

Institute of Intelligence Science and Technology, Computer and Information College, Hohai University, Nanjing 211100, China

Received:2015-09-21 Online:2016-03-02 Published:2016-03-03

摘要/Abstract

摘要： 根据Madaline网络工作原理,针对其网络特点和现有算法中存在的缺点，包括存在权值修改公式参数较多不容易协调,经验取值缺乏理论依据不够灵活，按照置信度原则进行翻转神经元会陷入“局部震荡”。提出改进的MRII学习算法,通过建立神经元敏感性替代置信度作为度量隐层神经元翻转的尺度，并采用感知机学习规则,减少权值调整次数，实验结果验证了该算法的优越性。

关键词: , 神经网络, Madaline, 学习算法, 敏感性, MRII

Abstract: In this paper, by analyzing the Madaline network work principle, according to the network characteristic and the shortcomings of the existing algorithms including the adjust weigh formula used in MRII algorithm with many parameters, but most of which come from experience in practice without theoretical reason, it made networks fall into “local recycle” with turning over the neuron based on confidence principle. We present an improved learning algorithm based on MRII, by establishing neuron’s sensitivity as a tool for measuring the turn of each hidden neuron, which reduces the number of weight adjustment. Some experimental results verify the effectiveness of the proposed algorithm.

Key words: neural network, Madaline, learning algorithm, sensitivity, MRII

郑星，李宗祥. 基于改进MRII的Madaline网络学习算法[J]. 计算机与现代化, doi: 10.3969/j.issn.1006-2475.2016.02.011.

ZHENG Xing, LI Zong-xiang. A Madaline Network Learning Algorithm Based on Improved MRII[J]. Computer and Modernization, doi: 10.3969/j.issn.1006-2475.2016.02.011.

参考文献

［1］胡守仁. 神经网络导论［M］. 长沙:国防科技大学出版社, 1993.
［2］张立明. 人工神经网络的模型及其应用［M］. 上海:复旦大学出版社, 1993.
［3］ Winter R, Widrow B. Madaline Rule II: A training algorithm for neural networks［C］// IEEE International Conference on Neural Networks. 1988: 401-408.
［4］ Winter R. Madaline Rule II: A new method for training networks for Adalines［D］. PhD Thesis, Department of Electrical Engineering, Standford University, 1989.
［5］张军英,许进. 二进前向人工神经网络理论与应用［M］. 西安：西安电子科技大学出版社, 2001.
［6］ Wang Yingfeng, Zeng Xiaoqin, Yeung Daniel So, et al. Computation of Madalines’ sensitivity to input and weight perturbations［J］. Neural Computation, 2006,18(11):2854-2877.
［7］ Zhang Chao, Yang Jie, Wu Wei. Binary higher order neural networks for realizing Boolean functions［J］. IEEE Transactions on Neural Networks, 2011,22(5):701-703.
［8］ Zeng Xiaoqin, Yeung D S. Hidden neuron pruning of multilayer perceptrons using a quantied sensitivity measure［J］. Neurocomputing-IJON, 2006,69(7-9):825-837.
［9］ Yeung D S, Ng W W Y, Wang Defeng, et al. Localized generalization error model and its application to architecture selection for radial basis function neural network［J］. IEEE Transactions on Neural Networks, 2007,18(5):1294-1305.
［10］Zhong Shuiming, Zeng Xiaoqin, Wu shengli, et al. Sensitivity-based adapitive learning rules for binary feedforward neural networks［J］. IEEE transactions on neural networks and learning systems, 2012,23(3):480-491.
［11］Yang S, Ho C, Siu S. Computing and analyzing the sensitivity of MLP due to to the errors of the i.i.d inputs and weights based on CLT［J］. IEEE Transactions on Neural Networks, 2010,21(12):1882-1891.
［12］Zeng Xiaoqin, Wang Yingfeng, Zhang Kang. Computation of Adalines’ sensitivity to weight perturbation［J］. IEEE Transactions on Neural Networks, 2006,17(2):515-519.
［13］Beak S M, Park J W. Hessian matrix estimation in hybrid systems based on an embedded FFNN［J］. IEEE Transactions on Neural Networks, 2010,21(10):1533-1542.
［14］哈根. 神经网络设计［M］. 戴葵译. 北京：机械工业出版社, 2002.

[1]	宋涛涛, 李艳萍, 李洪港, 韩春雪. 基于改进变结构趋近律的机械臂滑模控制系统[J]. 计算机与现代化, 2023, 0(12): 14-18.
[2]	周成诚, 曾庆军, 杨康, 胡家铭, 韩春伟. 基于高效通道注意力模块的运动想象脑电识别[J]. 计算机与现代化, 2023, 0(12): 19-23.
[3]	王宇航, 董宝良, 公超, 尚真真, 姚康宁. 基于意图识别的空中群目标动态威胁评估[J]. 计算机与现代化, 2023, 0(12): 100-104.
[4]	谷明轩, 范冰冰. 基于多模态特征融合的抑郁症识别[J]. 计算机与现代化, 2023, 0(10): 17-22.
[5]	陈俊义. 基于图节点动静态特征的健康事件预测模型[J]. 计算机与现代化, 2023, 0(10): 39-44.
[6]	黎世达, 项剑文. 一种提高图像识别模型鲁棒性的弱化强化方法[J]. 计算机与现代化, 2023, 0(10): 70-76.
[7]	陈子健, 段春红. 面向在线学习情境的认知情绪面部表情识别[J]. 计算机与现代化, 2023, 0(10): 92-98.
[8]	刘付琪, 张达, 宋建华, 王海东. 基于CNN-BiLSTM的液压系统故障诊断[J]. 计算机与现代化, 2023, 0(09): 10-19.
[9]	吴甜, 刘海华, 童顺延. 基于深度反馈的卷积神经网络的图像分类[J]. 计算机与现代化, 2023, 0(09): 82-86.
[10]	毛明扬, 徐胜超. 面向粒子群优化BP神经网络的粗糙集连续属性离散化算法[J]. 计算机与现代化, 2023, 0(09): 115-119.
[11]	孙子雨, 任燃, 魏曦哲. 基于DTW-TCN的股票分类及预测研究[J]. 计算机与现代化, 2023, 0(08): 31-37.
[12]	曾丽丽, 汤华贝, 牛艺晓, 孟凡月. 基于LSTM堆叠残差网络的岩相识别方法[J]. 计算机与现代化, 2023, 0(08): 38-43.
[13]	江蕾, 唐建, 杨超越, 吕婷婷. 基于CWGAN-GP与CNN的轴承故障诊断方法[J]. 计算机与现代化, 2023, 0(07): 1-6.
[14]	崔少国, 张岗, 王奥迪. 基于感知注意力的深度交叉网络推荐模型[J]. 计算机与现代化, 2023, 0(07): 54-60.
[15]	许叶彤, 耿信哲, 赵伟强, 张月, 宁海龙, 雷涛. 基于CNN-Transformer混合结构的遥感影像变化检测模型[J]. 计算机与现代化, 2023, 0(07): 79-85.

基于改进MRII的Madaline网络学习算法

A Madaline Network Learning Algorithm Based on Improved MRII

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价