Author(s) :

Dr.M.Deepa, M.P.Venkat Vijay, S. SriRanjani , V. Sowmiya, V. Ramya.

Volume/Issue :

Abstract :

Systems for detecting network intrusions that are based on anomalies are very important. This research proposes robust machine learning and deep learning models for classifying different forms of network intrusions and attacks. The 49-feature UNSW-NB15 dataset has been used in experiments by suggested models for nine distinct assault samples. Among the ensemble models, the Decision Tree classifier yielded the highest accuracy of 99.05%, followed by Random forest (98.96%), Adaboost (97.87%), and XGBoost (98.08%).The K-Nearest Neighbour classifier was trained for a range of K values, with K=7 yielding the best results and an accuracy of 95.58%. For binary classification, a Deep Learning model with two dense layers activated by ReLU and a third dense layer activated by Sigmoid was created. It yielded good accuracy of 98.44% when used with the ADAM optimizer and an 80:20 Train-Test Split Ratio. XGBoost detects network attack exploits with 95% accuracy, Random Forest detects fuzzer attacks with 90% accuracy, Random Forest detects generic assaults with 99% accuracy, and Decision Trees detects reconnaissance attacks with 79% accuracy. Detecting network attacks requires no feature selection because all features are powerful and important.

No. of Downloads :

0