|
|||||
|
|
| 基于最小二乘支持向量机的色谱指纹图谱预测银杏叶总抗氧化活性 | |
| 引用本文： | 姚卫峰,胡育筑,牟玲丽,余伯阳.基于最小二乘支持向量机的色谱指纹图谱预测银杏叶总抗氧化活性[J].分析化学,2009,37(3). |
| 作者姓名： | 姚卫峰 胡育筑 牟玲丽 余伯阳 |
| 作者单位： | 1. 南京中医药大学药物分析教研室,南京,210046;中国药科大学分析化学教研室,药物质量与安全预警教育部重点实验室,南京,210009 2. 中国药科大学分析化学教研室,药物质量与安全预警教育部重点实验室,南京,210009 3. 中国药科大学中药复方研究室,现代中药教育部重点实验室,南京,210009 |
| 摘 要： | 在色谱图基线校正和色谱峰匹配基础上,提出以40个银杏叶提取物HPLC指纹图谱的色谱图轮廓作为输入,相应的提取物总抗氧化活性作为输出,建立最小二乘支持向量机回归模型,并对包含10个样本的测试集进行了预测.最小二乘支持向量机的测试集预测误差均方根(RMSEP)为0.0230,预测结果优于目前普遍使用的误差反向传播神经网络和偏最小二乘回归.与采用色谱峰面积为分析变量的模型预测结果比较表明:采用消除干扰后的色谱图全谱轮廓保留了样本的全部信息,预测结果更好 |
| 关 键 词： | 中药色谱指纹图谱 色谱图轮廓 基线校正 峰匹配 最小二乘支持向量机 总抗氧化活性 |
Prediction of Total Ginkgo Biloba Leaves Antioxidant Capacity from Chromatographic Fingerprints by Least Squares-Support Vector Machines |
|
| YAO Wei-Feng,HU Yu-Zhu,MOU Ling-Li,YU Bo-Yang.Prediction of Total Ginkgo Biloba Leaves Antioxidant Capacity from Chromatographic Fingerprints by Least Squares-Support Vector Machines[J].Chinese Journal of Analytical Chemistry,2009,37(3). | |
| Authors: | YAO Wei-Feng HU Yu-Zhu MOU Ling-Li YU Bo-Yang |
| Abstract: | With the entire high performance liquid chromatographic profiles after baseline correction and peak alignment,a predictive least squares support vector machines model was built for the prediction of Ginkgo biloba leaves total antioxidant capacity using a training set including 40 samples.Furthermore,the total antioxidant capacity of a test set with 10 samples was predicted.The root mean squared error of prediction(RMSEP) of least squares-support vector machines(LS-SVM) was 0.0230,which was performed slightl... |
| Keywords: | Chromatographic fingerprints of traditional Chinese medicine chromatographic profiles baseline correction peak alignment least squares-support vector machines total antioxidant capacity |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |