- Purvit Vashishtha
- Arun Kothari
- Prabhishek Singh
- Manoj Diwakar
- Neeraj Kumar Pandey
The application of machine learning and also deep learning strategies in the field of cyber safety and security is a lot more popular than in the past. From IP (Internet Protocol) web traffic classification, a filtering system for harmful web traffic for breach discovery, Machine learning is among the appealing solution that can be reliable against zero-day dangers. This paper is a concentrated study of machine learning and also its application to the cyber division for breach discovery, web traffic classification as well and also applications such as e-mail filtering systems. Ensemble learning is being implemented in this paper which has two parts namely bagging and boosting. It is basically focused on combining multiple models in order to solve a particular problem whether it be regression or classification, which in this problem statement is classification. Applicability of these methods in different cybersecurity tasks such as breach identification, recognition of viruses, phishing, forecasting cyberattacks, e.g., denial of service, fraud recognition or cyber anomalies, and many more are going to be reviewed in this paper additionally. On the whole, the objective of this paper is to have an introduction of exactly how ensemble learning can be shown reliable and also a brilliant aspect for the academic community as well as experts in the cyber industries for identifying attacks.
- Machine Learning
- Ensemble Methods
- Bagging
- Boosting
- Evaluation Methods
NSL-KDD dataset from kaggle is used in this research paper. NSL-KDD is a data set suggested to solve some of the inherent problems with the KDD'99 dataset. Although, this new version of the KDD data set still suffers from some of the problems, it still can be applied as an effective benchmark data set to compare different intrusion detection methods.