Real-time Intrusion Detection System implementing Machine Learning. We combine Supervised Learning (RF) for detecting known attacks from CICIDS 2018 & SCVIC-APT datasets, and Unsupervised Learning (AE) for anomaly detection.

Network data classifier based on the recurrent neural network.

Deep Model Intrusion Detection (IDS) Evaluation of NSL KDD and CIC IDS 2018 datasets.

DNN-Ensemble IDS is a machine learning based classification model for intrusion detection exploiting ensembles of classifiers.

This project aims to identify and classify the anomalies captured in network traffic using different machine learning strategies. After the reults are given, I compared the results of two classical approaches for supervised learning: RandomForest and SVM on a large public combined dataset made from CICIDS2017 dataset and CICIDS2018.