A Deep Learning- Based Intrusion Detection and PreventionSystem For Detecting and Preventing Denial-of-Service Attacks

Abstract Denial-of-Service (DoS) attacks pose a significant threat by overwhelming network resources and rendering services inaccessible to legitimate users, while traditional detection systems often fail to keep pace with evolving attack patterns. To overcome this limitation, the proposed project applies machine learning techniques using the dataset dataset_sdn.csv to train and evaluate supervised models such as Logistic Regression, K-Nearest Neighbors, Support Vector Machine (SVM), Decision Tree, Random Forest, and Gradient Boosting, comparing their performance in terms of accuracy, precision, recall, and computational efficiency. To enhance detection further, a deep learning-based Long Short-Term Memory (LSTM) neural network is integrated, leveraging its ability to capture sequential dependencies and time-series anomalies that are critical in identifying sophisticated DoS attack behaviors. This combined approach not only improves detection accuracy but also supports real-time adaptation, with the final aim of embedding the framework into an Intrusion Detection and Prevention System (IDPS) for proactive threat mitigation. By uniting machine learning, deep learning, and real-time deployment, the project offers a robust and resilient defense mechanism against DoS attacks. Keywords: Denial-of-Service, Machine Learning, Deep Learning, LSTM, Intrusion Detection.