APPLICATION OF MACHINE LEARNING FOR INTRUSION DETECTION SYSTEM

Due to Covid-19 pandemic, the most of the organizations have permitted their employees to work from home. Also, it is every essential to have security at the highest level so that information will flow in the safe and trusted environment between the different organizations. There is always threat of misuses and different intrusions for communication of the data securely over the internet. As more and more people are using online transactions for the different purposes, it is found that the cyber attackers have become more active. Three in four organizations have faced the different cyber-attacks in the year 2020. So, the detection of intrusion is very important. The paper introduces the intrusion detection system and describes its classification. It discusses the different contributions to the literature in literature review section. The paper discusses the application of the different feature selection techniques for reducing the number of features, use of the different classification algorithms for the intrusion detection and it shows how machine learning is used effectively. KDD99 benchmark dataset was used to implement and measure the performance of the system and good results are obtained and the performance of the different classifier algorithms was compared. Tree based classifiers such as J48 and ensemble techniques such as random forest give the best performance on KDD99 dataset.