基于Stacking集成学习的链路质量评估模型  

Link Quality Estimation Model Based on Stacking Ensemble Learning

作　　者：胡余强 HU Yuqiang(Nanchang Hangkong University,Nanchang 330063)

机构地区：[1]南昌航空大学,南昌330063

出　　处：《计算机与数字工程》2025年第2期510-516,共7页Computer & Digital Engineering

摘　　要：为了提高不同实验场景中无线传感网络的链路质量评估效果,论文通过分析单分类器与集成分类器分别在链路质量样本数据中分类效果选择综合效果好的基学习器,然后在基学习器的基础上选择元学习器,实现构建一种基于Stacking集成学习的链路质量评估模型。实验表明,该评估模型在准确率、精确率、召回率、F1-score评估指标上有更好的效果。In order to improve the link quality estimation effect of wireless sensor networks in different experimental scenari⁃os,the paper selects a basic learner with a good comprehensive effect by analyzing the classification effect of the single classifier and the ensemble classifier in the link quality sample data,and then uses the basic learning method in the basic learning method.The meta-learner is selected on the basis of the device,and a link quality evaluation model based on Stacking ensemble learning is constructed.Experiments show that the evaluation model has better performance in terms of accuracy,precision,recall and F1-score estimation evaluation indicators effect.

关 键 词：无线传感网络 链路质量评估 Stacking集成学习 

分 类 号：TP393.17[自动化与计算机技术—计算机应用技术]