Network traffic classification algorithm based on improved ACO-SVR

This study proposes an improved ACO-SVR algorithm for network traffic classification. The improved ACO algorithm is used to optimize the parameters of SVR model. ACO algorithm dynamically adjusts the parameter search range according to the number of iterations to improve the search accuracy and convergence speed. The heuristic information algorithm is based on the correlation calculation between features and target variables to guide ant search. Experimental results show that this method has significant advantages for classification accuracy and efficiency.