一种基于层级分类策略的复杂模式识别方法  

Method of complicated pattern recognition based on hierarchical classification strategy

作　　者：赵漫丹 郝向阳[1] 张振杰[1] 

机构地区：[1]解放军信息工程大学导航与空天目标工程学院,郑州450001

出　　处：《计算机应用研究》2016年第8期2320-2323,2330,共5页Application Research of Computers

摘　　要：针对大量复杂而亟待分类的模式类型,提出了一种基于层级分类策略的复杂模式识别方法。该方法通过选择合适的分层标志,从多侧面反映被分类对象的属性,降低了分类的复杂度,提高了分类效率。建立了层级分类策略模型以及关键性步骤,并将其应用于利用表面肌电信号识别手部动作的实验验证,通过对比是否采用此策略,进而说明了基于层级分类策略在面对复杂模式识别的有效性,为该方法的推广使用奠定了基础。For the large number of complex and urgent need for recognizing pattern types, this paper proposed a method of complicated pattern recognition based on hierarchical classification strategy. The method must determine the hierarchical sign. Reacting the attribute of the object from the muhi sides. From this,it reduced the complexity of the classification, and improved the classification efficiency. It established the hierarchical classification strategy model and the key steps. Then, the experiment about recognizing hand movements by using of surface electromyography signal, it can compare the recognition effect whether to use of this strategy. Thus,it verifies the effectiveness of this strategy, and also becomes a template to promote this method.

关 键 词：层级分类 模式识别 肌电信号 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]