Application of Machine Learning in Cigarette Quality Evaluation Method

In the evaluation of cigarette quality, it usually depends on the sense organs of evaluation experts. But due to the influence of many subjective and objective factors, the accuracy of evaluation results is often difficult to guarantee. Therefore, this paper proposes a method of cigarette quality evaluation based on SVM and BP neural network. The experimental results show that the method can establish a nonlinear mapping relationship between the measured values of cigarette chemical components and the evaluation values of expert quality, and reflect the preference information and reasoning mechanism of evaluation experts. So the model is an objective and reliable method for evaluating the internal quality of tobacco after knowing the chemical composition test data of tobacco.