摘要
研究人们交互学习过程,建立神经网络模型,利用系统仿真技术探索发现和掌握知识的基本规律,有利于促进和提高人们的智慧。用博弈理论框架模拟交互学习过程;用人工神经网络完成学习功能;建立交互学习神经网络模型。在此基础上,研究交互学习模型特性,求解了混合策略博弈过程的均衡预测问题。在神经网络学习过程中,效仿人们回顾对比的学习方法,更新神经网络连接权值。利用建立的交互学习神经网络模型,进行系统仿真研究。仿真结果表明,该模型不仅能很好的模拟人类交互与竞争学习过程,还能对博弈过程的均衡状态做出有效预测。
Exploring the basic hales of human discovering and grasping knowledge, establishing its mathematical model and utilizing system simulation technology to promote the process of discovering knowledge are of vitally important factors for accelerating human intelligence. Specially, in an increasingly complex, crowded and competitive world, these factors are extremely significance for a nation to stand among worldly nationalities. Along this direction the game process model was put forward to combine with artificial neural networks to model the interactive learning process. The repeated game paradigm was used to model the competitive process; and neural networks were utilized as a learning model Then an interactive and competitive model was established by combining the game paradigm with the neural networks. Using the established interactively competitive learning model, system simulation studies were conducted. The simulation results indicate that the model can be used to predict the equilibrium of mixed strategy repeated game process. The model can also be used in the investigation of human interactive and competitive learning behavior.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2009年第17期5314-5317,共4页
Journal of System Simulation
基金
交通部交通应用基础研究项目(200332922505)
关键词
学习模型
神经网路
系统仿真
博弈论
learning model
neural networks
system simulation
game theory
作者简介
任光（1952-），男，辽宁朝阳人，博士，教授，博导，研究方向为轮机自动化与控制，复杂系统建模。
