摘要
神经网络常备用于解决拟合回归问题以及植物种类的分类识别。常用的植物种类分类识别方法是分类检索表,但费时费力,这种分类检索表法由于直观而被人们加以使用。根据GRNN(广义回归神经网络)和PNN(概率神经网络)理论,在MATLAB中对鸢尾花的相关属性建模,根据收集到的鸢尾花种类数据进行了分析与仿真,大大回避了分类检索表法的缺点。仿真结果表明,GRNN神经网络在鸢尾花种类识别中比PNN精准度高,对于植物种类的分类识别应用前景更加广泛。
Generalized regression neural network is a kind of neural network with mentors learning.It is used in many fields but is mostly used to solve regression problems and identify the classification of plant species.The commonly way of classification and identification of plant species is the classification and retrieval table method.It is time-consuming and laborious,but this classification retrieval table method is used by people because of intuitiveness.However,according to the GRNN neural network and PNN neural network theory,the related attributes of Iris flower were modeled in MATLAB,and based on the collection of Iris flower species data,the analysis and simulation were carried out,which greatly avoided the disadvantages of the classification retrieval table method.The simulation results show that in the identification of iris species,the GRNN neural network is more accurate than the PNN,so the GRNN neural network has more extensive application prospects for the classification and identification of plant species.
作者
康彩丽
KANG Caili(Hunyuan Normal School Attach to Datong University,Hunyuan 037400,China)
出处
《忻州师范学院学报》
2020年第2期17-21,共5页
Journal of Xinzhou Teachers University
关键词
GRNN神经网络
鸢尾花
种类识别
植物分类
GRNN neural network
iris flower
species identification
plant classification