基于广义回归神经网络的薄荷属植物的分类识别  

Classification and Recognition of Mentha Based On the Generalized Regression Neural Network

作　　者：黄劲潮[1] 

机构地区：[1]龙岩学院继续教育学院,福建龙岩364000

出　　处：《唐山师范学院学报》2015年第2期45-48,64,共5页Journal of Tangshan Normal University

基　　金：龙岩学院校立服务海西项目(lyxy2011057)

摘　　要：薄荷属植物易杂交以致多型性现象严重,对于薄荷属植物的分类主要采用分类检索表法或数量分类法等,但这些方法花费的时间与物力较大。选取了薄荷属植物类别区分的一些强相关属性作为训练组,构建了广义回归神经网络来对薄荷属植物进行分类、识别,在保证准确度的同时也兼顾了识别的速度。同时,该模型也可用来识别薄荷属植物的某种属性是否与类别区分强相关。Mentha is easily to hybrid that polymorphism phenomenon is serious, therefore, the classification of Mentha mainly uses classification method or the number of classification and so on, but it spends a lot of time and effort of these methods. This paper selects some strong related attributes of plant category distinctions of Mentha as the training set, constructs the Generalized Regression Neural Network with mentors to classify and recognize Mentha, and ensures the accuracy and the recognition speed. At the same time, the model can also be used to tell whether some certain attribute of Mint is strongly related with classification recognition or not.

关 键 词：GRNN 薄荷属 分类 识别 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]