Archives

Boosting Using Ensemble for Improving the Accuracy of E-mail Spam Classification


Prakash V. Parande and Dr.M.K. Banga
Abstract

Boosting is one of the most effective ensemble methods designed for classification. Boosting enhances the capabilities of a classifier in large data set. Classification of e-mails as spam is a crucial problem in internet electronic communication, with bulk e-mails in the inbox and is a challenging task to filter it out. This paper, proposes Boosting for improving accuracy of e-mail spam classification. The boosted decision tree algorithm is used to improve the accuracy of e-mail classification. A mathematical model to boost the accuracy of e-mail classifier and an algorithm to prove the training an test set split ratio are developed to show the improvement in accuracy during classification. The experimental results show that the proposed algorithm has better accuracy by around 4% compared to the existing decision tree algorithm.

Volume 11 | 01-Special Issue

Pages: 973-981