摘要
针对智慧校园可视化平台进行系统设计,提出用户活跃度和轨迹相似度算法,构建用户画像模型,最后对提出的智慧校园可视化平台算法进行测试分析。结果显示,优化后的用户相似度算法更能为精确,能够较好的度量移动轨迹相似性。提出的特征选择算法的分类结果准确度较高,优于其他算法,集成模型的标签预测准确率明显优于简单分类器,二级融合模型的准确率在一级Stacking的基础上又提升了3%。
This study designs the system for the intelligent campus visualization platform,proposes the algorithm of user activity and trajectory similarity,constructs the user profile model,and finally tests and analyzes the proposed algorithm of smart campus visualization platform.The results show that the optimized user similarity algorithm is more accurate and can better measure the similarity of moving tracks,the classification accuracy of the feature selection algorithm proposed in this study is higher than other algorithms,the accuracy of the integrated model is significantly better than that of the simple classifier,the accuracy of the two-level fusion model is improved by 3%on the basis of the first-order stacking.
作者
颜杰森
YAN Jiesen(Information Center of Party and Government Office,Quanzhou Vocational College of Arts and Crafts,Quanzhou,Fujian 362500,China)
出处
《武夷学院学报》
2022年第3期41-46,共6页
Journal of Wuyi University
基金
福建省教育科学“十三五”规划课题《基于物联网的智慧校园可视化平台研究》(FJJKG20-066)。
关键词
物联网
智慧校园
可视化
活跃度
轨迹相似度
用户画像
internet of things
smart campus
visualization
activity
trajectory similarity
user profile
