Triple Network Intrusion Detection Model Using Machine Learning

Abstract

Due to the enormous amounts of data and their gradual growth, systems for Big Data analysis and information security have recently changed in terms of their relevance. A framework called an interruption location framework (IDS) screens and examinations information to track down any interruptions into a framework or organization. Due to the network's rapid data generation, volume, and variety, traditional approaches are no longer practical. extremely difficult to identify attacks. IDS utilizes huge information ways to deal with handle large information for exact and viable information investigation. The random forest model suggested this for intrusion detection. An intrusion detection model was constructed on the Apache Spark Big Data platform with the help of a Random Forest classifier and ChiSqSelector. The model was trained and evaluated using KDD99. In the experiment, we contrasted Linear Discriminant Analysis, Decision Tree, Random Forest, and Logistic Regression. The experiment's findings demonstrated that the Random Forest model performs well, requires less time to train, and is effective when used with large amounts of data.