| 二进制互连神经网络的遗传演化 |
| |
| 引用本文： | 吕阳鹏,李大字,靳其兵,谭天伟. 二进制互连神经网络的遗传演化[J]. 北京化工大学学报(自然科学版), 2010, 37(4): 126-130. DOI: 10.3969/j.issn.1671-4628.2010.04.026 |
| |
| 作者姓名： | 吕阳鹏 李大字 靳其兵 谭天伟 |
| |
| 作者单位： | 北京化工大学 1 信息科学与技术学院; 2 生命学与技术学院， 北京 100029 |
| |
| 基金项目： | 国家"863"计划，国家"973"计划，北京市优秀人才资助项目 |
| |
| 摘 要： | 当前对于二进制神经网络的研究主要集中在前馈模型上,然而前馈模型不具备反馈连接,无法形成记忆结构,从而无法处理时序逻辑问题。提出了一种基于遗传算法的二进时序神经网络演化算法,并且通过双边沿触发计数器实验和迷宫路径覆盖实验验证了演化的二进制互连神经网络具备时序逻辑处理能力。
|
| 关 键 词： | 二进制神经网络 时序逻辑 遗传算法 迷宫路径覆盖 |
| 收稿时间： | 2009-12-12 |
A genetic algorithm-based binary neural network generator |
| |
| LV YangPeng,LI DaZi,JIN QiBing,TAN TianWei. A genetic algorithm-based binary neural network generator[J]. Journal of Beijing University of Chemical Technology, 2010, 37(4): 126-130. DOI: 10.3969/j.issn.1671-4628.2010.04.026 |
| |
| Authors: | LV YangPeng LI DaZi JIN QiBing TAN TianWei |
| |
| Affiliation: | 1. College of Information Science and Technology; 2. College of Life Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China |
| |
| Abstract: | Current research on Binary Neural Networks is focused on the Feed Forward Model. However,the Forward Model cannot form backward links,which are a necessary part of memory structure and,therefore,it cannot be used to solve sequential logic problems. This paper presents a Genetic Algorithm-based Binary Neural Network Generator,whose capability of creating sequential logic neural networks is verified by two experiments:a double-edge triggered counter experiment and a maze path covering experiment. |
| |
| Keywords: | binary neural network sequential logic genetic algorithm maze path covering |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《北京化工大学学报(自然科学版)》浏览原始摘要信息 |
|
点击此处可从《北京化工大学学报(自然科学版)》下载全文 |
