为了提高近红外光谱快速检测烟草尼古丁含量的精度和稳定性，利用近红外光谱结合遗传算法最小二乘支持向量回归(GA-LSSVR)建立了回归预测模型。在LSSVR模型建立过程中，采用遗传算法对LSSVR参数进行自动优化。相比于利用常规最小二乘支持向量机和遗传偏最小二乘法等建立的回归预测模型，GA-LSSVR法建立的回归预测模型泛化能力更强,预测效果更好，验证集相关系数R2为0.9766，预测均方根误差为0.1065。研究结果表明，GA-LSSVR是一种快速准确的建模方法，为烟草尼古丁含量的近红外测定和近红外光谱数据的处理提供了新的方法与途径。

In order to improve the detecting precision and robustness in determination of the content of nicotine in tobacco by near infrared spectroscopy, a predictive model is established by genetic algorithm combined least square support vector regression (GA-LSSVR). The practical use of support vector machine is limited because of its set of parameters to be defined by the user. For this reason, a genetic algorithm is utilized to approach LSSVR parameter optimization in the calibration model. To highlight the superiority of GA-LSSVR algorithm, it is compared with traditional LSSVR and genetic algorithm-partial least square. The correlation coefficient R2 and root mean square error of prediction (RMSEP) for the test set are used as evaluation parameters for the model. The optimal model is obtained with R2 of 0.9766 and RMSEP of 0.1065. Generally, in the context of performance and robustness, the results demonstrate that GA-LSSVR is a good method for the analysis and modelling of near infrared data.