A Novel Model for Anomaly Detection in Network Traffic Based on Support Vector Machine and Clustering

Algorithm 3

The training algorithm for SVM-C.

 Input: , k, d, the maximum number of iterations , and initial

 Output:

 Step 1. Let .

 Step 2. Apply the feature extraction method in Section 3.2 to obtain the dataset and divide it into the training set and test set.

 Step 3. Solve subproblem (2) using the PBB method and obtain . Compute the accuracy on the current testing set.

 Step 4. Solve subproblem (6) using the PBB method and obtain .

 Step 5. If , return and corresponding to the maximum accuracy and terminate the algorithm. Otherwise, let and return to Step 2.