摘要
文章先提出了一种基于窄带物联网(Narrow Band Internet of Things,NB-IoT)通信的在智能农业领域中的典型应用系统,给出了系统的设计框图。由于统计学习方法在物联网领域中经常结合使用,文章提出了一种结合逻辑斯谛回归(Logistic Regression,LR)和支持向量机(Support Vector Machine,SVM)的混合模型。实验表明,该混合模型比单独使用LR模型或者SVM模型的分类准确率更高,并且使用子集进行训练的特点能够减少SVM的迭代次数,提升了算法的性能。
In this paper,a typical application system in the field of intelligent agriculture based on Narrow Band Internet of Things(NBIoT)communication is proposed,and the design flow chart of the system is also given.As statistical learning methods are often used in combination in the field of Internet of Things,a hybrid model combining Logistic Regression(LR)and Support Vector Machine(SVM)is proposed.Experiments show that the hybrid model has higher classification accuracy than using the LR model or the SVM model alone,and the feature of using subsets for training can reduce the number of iterations of the SVM and improve the performance of the algorithm.
作者
林启明
Lin Qiming(School of Electronic and Optical Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210023,China)
出处
《无线互联科技》
2020年第23期32-35,共4页
Wireless Internet Technology
关键词
窄带物联网
支持向量机
逻辑斯谛回归
混合模型
Narrow Band Internet of Things
Support Vector Machine
Logistic Regression
hybrid model