Uncertainty quantification method for classified prediction of encrypted network traffic

The invention relates to an uncertainty quantification method for classified prediction of encrypted network traffic, and belongs to the field of network security. The method comprises the steps of model training prediction, Gaussian kernel density estimation, KL divergence calculation, OOD score calculation, hypothesis verification and the like. According to the method, uncertainty quantitative analysis can be carried out on the prediction result of the existing encrypted network traffic classification model, the credibility of the prediction result is detected, and performance evaluation is carried out on the existing model. In the prediction process, the method can quantify the uncertainty of the prediction result, calculate the OOD score through Gaussian kernel density estimation and KL divergence, measure the similarity and difference between samples inside and outside the distribution, and help the system to identify and adapt to new data, thereby improving the expansibility and robustness of the model.