中文摘要：

基于烟叶化学数据建立烤烟香型分类模型，然后对各模型进行筛选比较选出最优模型。首先对142个烤烟烟叶样品中的9类成分的63个指标采用行业标准进行检测，然后采用逐步回归法筛选出19个烟叶化学成分．依据这19个指标采用线性判别分析法、Logistie回归、高斯混合模型、分类树、K最邻近法、人工神经网络和支持向量机七种方法进行建模。通过对不同方法建立的模型采用100次随机抽取训练集样本和测试样本计算错误分类率．选择错误分类率较低的模型作为优选模型。经比较发现，线性判别法和高斯混合模型建立的两种香型函数能较好地对未知样品的香型进行正确分类．且效果较好。筛选出的两种优选模型对于烤烟香型分类研究具有一定的应用价值。

英文摘要：

Based on the chemical components of tobacco leaves, the classification models of tobacco flavor were established. All models were compared to select the optimal model. 63 components of 9 kinds of 142 tobacco leaves were detected by to- bacco industry standards. 19 chemical components were selected by stepwise regression method. Seven methods including dis- criminate analysis, Logistic regression, Gauss mixture model, classification tree, K nearest neighbor method, artificial neural network and support vector machine were used to establish the models based on the 19 index. 100 randomly selected samples were used as the training sets and test samples to calculate the error classification rate through the establishment of the dif- ferent methods of models. The model was the preferred model with classification error rate lower than others. By comparision, two kinds of flavor function model （linear discriminate method and Gauss mixed） were better to unknown sample types. Two kinds of optimization models had a certain application value for classifying tobacco flavor.