Enhanced Pre-Processing Techniques based Intrusion Detection Systems using Machine Learning

Abstract

These days, intrusion detection system (IDS) is the most arising pattern in our general public. This basically screen network traffic and will alarm the organization chairman of any unordinary action. IDS System work by one or the other searching for marks of known assaults or deviations of typical movement. While there are a few detriments of IDS, for example, low recognition rate and high bogus caution rate. In this paper a mixture IDS (HIDS) strategy dependent on support vector machine (SVM) and evidence theory (ET) has been proposed too different assault recognition method to limit the low bogus alert rate and improve exactness.