近红外光谱技术是一种简单,快速,无损,价格低廉的方法,可以进行多组分同时分析.支持向量机基于结构风险最小化原理替代了传统方法中的的经验风险最小化原理,使得它具有更好的泛化能力,在许多领域中的应用取得了成功.在这篇文章中,我们把连续小波变化技术结合支持向量机用于近红外光谱分析,结果显示,连续小波变换-支持向量回归模型具有更好的预测精度. Near Infrared Spectroscopy is a simple, rapid, non-destructive, low-cost method for simultaneous multi-component analysis.VM based on the principle of structural risk minimization replaces the principle of empirical risk minimization in traditional methods, And has been successfully applied in many fields.In this paper, we use continuous wavelet transform technology and support vector machine for near-infrared spectroscopy, the results show that the continuous wavelet transform - support vector Regression models have better prediction accuracy.