Enhancing Network Security: Leveraging Machine Learning for Intrusion Detection

Abstract

In order to precisely identify and categorise different kinds of network attacks, this study focuses on applying machine learning approaches for network intrusion detection. Data collection, preprocessing, feature scaling, model definition, feature selection, and assessment metrics are all part of the methodology. Different machine learning models, including Decision Tree Classifier and Random Forest Classifier, are considered, along with the use of all features or some part of features for each attack category. Evaluation is performed using K-fold cross-validation, with metrics such as accuracy, precision, recall, and F1-score analysed. Results indicate the efficiency of Random Forest Classifier in handling high-dimensional datasets and improving detection accuracy, making it a superior choice for network intrusion detection tasks.