Long Short-Term Memory and Synthetic Minority Over Sampling Technique-Based Network Traffic Classification

The swift expansion of network data and the growing complexity of cyberattacks demand sophisticated techniques for precise and prompt categorization of network traffic. The capacity to accurately categorize network traffic as BENIGN or malicious/attack is critical in the field of cybersecurity. The goal of this study was to create a system for classifying network traffic by combining the Synthetic Minority Oversampling Technique (SMOTE) with Long Short-Term Memory (LSTM). Recurrent neural networks (RNNs) of the LSTM type are renowned for their capacity to capture temporal dependencies in sequential data. On the other hand, SMOTE is a technique that balances the distribution of classes by using the oversample of the minority class (malicious traffic) and applies SMOTE only to the training data, preventing data leakage into the test set.