Comparative Study on Different Classification Techniques for Spam Dataset

Abstract

Nowadays, people and companies use emails for information exchange, email messages, and etc., because they are the fastest and the cheapest way. The main problem that faces email messages is the undesirable emails which known as spams. Spams may cause overflow the internet with considerable copies of the same message or carry malicious content that harms user system and reduce the performance. The purpose of this work is to make a comparative study of several classification techniques on the basis of their performance parameters using spam dataset. The performance of the different classifiers is measured with different ratio of the testing and training dataset. Also, the performance of the classifiers is calculated with and without low variance filter. By applying the low variance filter the accuracy of the KNN classifier is enhanced with about 9% while the accuracy of the other classifier is decreased.