摘要
基于特征选择优化方法有利于检测挥发性有机化合物的毒性.采用威尔克斯Λ统计量筛选关键特征,结合动态模式检测模型,可以提升分类与浓度预测性能.结果表明,该方法的分类准确率高,浓度预测均方根误差低,在复杂环境下具有一定有效性.
An optimized feature selection methodology proves advantageous for assessing the toxicity of volatile organic compounds(VOCs).By employing WilkesΛstatistic for key feature screening and integrating a dynamic pattern recognition model,this approach significantly enhances classification performance and concentration prediction accuracy.Experimental results demonstrate that the proposed method achieves high classification accuracy,low root mean square error(RMSE)in concentration forecasting,and maintains notable effectiveness under complex environmental conditions.
作者
廖洪燕
Liao Hongyan(Sichuan Preschool Educators College,Mianyang 621000,China)
出处
《许昌学院学报》
2025年第5期42-46,共5页
Journal of Xuchang University
关键词
挥发性有机物
特征选择
检测
volatile organic compounds(VOCs)
feature selection
detection
