An Estimation of Machine Learning Approaches for Intrusion Detection System

IDS is a machine approach that can prevent or analyze network attacks. An attacker would concession the confidentiality, availability, and integrity of network resources. Due to these resources, there some sorts of violation exist. Thus, there arises a dire need to estimate the current scenario of harmful network attacks. One possible solution for protection over vulnerability is utilizing IDS system with machine learning techniques. Here, evaluation of machine learning techniques has been done using KddCup99 dataset. For this purpose, a correlation-based feature selection method based on ranking feature is used in order to get more accurate and efficient results followed by high detection rate. A performance matrix has also been developed for every classifier based on their false positive and false negative values.