节点文献
基于容器理论的模式识别算法
The Algorithm of Pattern Recognition Based on Container Theory
【摘要】 容器算法一般用于复杂系统的拟合和仿真,本文将其推广应用到模式分类问题之上,主要方法有拟合指纹比较法和训练法两种。文章在最后通过例子结合CUP分形特性探讨了不同时钟信号序列对演化序列的影响和不同区分度的现实意义。
【Abstract】 The container algorithm is generally used for the fitting and simulation for complex systems. This paper applies it to the pattern classification problem. The main methods are fingerprint of fitting comparison method and training method. In the end, the article discusses the fractal characteristics of pattern recognition by examples. and the different clock signal sequences on the evolutionary sequence and the practical significance of different degrees of discrimination are discussed.
【关键词】 CUP算法;
演化序列;
拟合指纹比较;
训练法;
【Key words】 CUP algorithm; evolutionary sequence; fingerprint of fitting; training method;
【Key words】 CUP algorithm; evolutionary sequence; fingerprint of fitting; training method;
【基金】 上海市科技创新行动计划项目资助(编号16511101200)
- 【文献出处】 自动化技术与应用 ,Techniques of Automation and Applications , 编辑部邮箱 ,2019年06期
- 【分类号】TP391.4
- 【被引频次】2
- 【下载频次】39