Identification of Distributed Denial of Service Attacks Employing L-BFGS Tuning
Vikas Singh, Vasudev Dehalwar, Jyoti Bharti · 2023
Internet speed and capacity are crucial components of any communication technology in the modern era. Customers should expect excellent speed and a lot of bandwidth. A DDoS (Distributed denial of service) assault, which is a well-known attack, is the biggest threat in communication. DDoS attacks limit the user’s speed and bandwidth. In order to produce this work, we used the updated dataset, CICDDoS2019, and undertook tests using a variety of machine learning methods. We investigated which features had the strongest relationships with the analysis’s projected classes. This research proposal offers a machine learning-based L-BFGS (Limited-memory Broyden-Fletcher-Goldfarb-Shanno) algorithm-based DDoS assault detection method. An HPC Python platform was used to carry out the simulation and experiment for the suggested research project. The accuracy, true and false negatives, along with true and false positives, confusion matrix structures, and other outcome measures are all evaluated in this study. The proposed method shows better results as compared to other previous methods. Index Terms—DDoS attack, MLP Classifier, LBFGS, Machine Learning.