Study of spam filter algorithm based on CS-SVM and Bagging
Bian Jirong · Ningxia Engineering Technology · 2008
According to the asymmetry of legitimate message and spam in the application of spam filtering,a spam filter algorithm is proposed on the basis of CS-SVM and Bagging.It tries to make the optimal classifier and improve the precision of spam filtering by defined every sample with different cost and by using the means of minimizing the misclassification cost.The data of the final simulation test can prove that the algorithm has high precision and efficiently lower error rate to a non-Spam.