论文部分内容阅读
用组合近红外光谱和化学计量学方法——相似分类法(SIMCA)和支持向量机(SVM)——来识别不同品牌烟草产品.通过一个案例研究发现,当训练样本较少时,SVM分类器优于SIMCA分类器;而当训练样本相对充足时,两类分类器精度无显著差异;SVM分类器对训练集组成依赖较小.结果表明:支持向量机组合近红外光谱是一个烟草质量控制中有价值的工具,可改变长期以来烟草产品的质量评估依赖于专家的感官这一传统方法,为评估结果的准确性和公正性提供有力手段.
Using a combination of near infrared spectroscopy and chemometrics methods SIMCA and SVM to identify different brands of tobacco products, a case study shows that when there are fewer training samples, the SVM classifier Which is better than that of SIMCA classifier.The accuracy of the two classifiers is not significantly different when the training samples are relatively abundant.The SVM classifier has less dependence on the training set.The results show that the support vector machine combination near infrared spectroscopy is a tobacco quality control Valuable tools that change the traditional way that quality assessments of tobacco products have long relied on the senses of experts provide a powerful tool for assessing the accuracy and fairness of results.