期刊论文

中文

武汉轻工大学学报

2095-7386

2021

40

6

41-46

10.3969/j.issn.2095-7386.2021.06.006

国家自然科学联合面上基金NSFC-CAAC（U1833119）；

本校为第一机构

BP神经网络是目前最为重要、应用最广的人工神经网络算法之一。详细阐述了BP神经网络结构及其工作流程。在此基础上,以Sklearn提供的鸢尾花数据集作为数据来源,设计了3层BP神经网络结构,实现了鸢尾花的分类。为了减少模型过拟合,在损失函数后加入了L1正则项。经训练和测试,BP神经网络能够有效地实现鸢尾花分类,成功率可达100%。通过加入L1正则项改进BP神经网络结构,可以大大减少训练时间,提高网络效率。实验表明,基于BP神经网络的鸢尾花的分类研究,具有实用性和有效性,对于植物的分类识別研究具有重要意义。

BP neural network is one of the most important and widely used artificial neural network algorithms.This paper expounds the structure and work flow of BP neural network in detail.On this basis,the iris data set provided by sklearn is used as the data source.A three-layer BP neural network structure is designed to realize iris classification.In order to reduce the over fitting of the model,L1 regularization is added to the loss function.After training and testing,BP neural network can effectively realize iris classification,and the success rate can reach 100%.By adding L1 regular term to improv...