Machine Learning Techniques for the Prediction of Cyber-Attacks

The aim of this paper to develop a machine-learning (ML) model for UNSW-NBI5 dataset cyberattack prediction. ML is a subset of AI which is a powerful tool for detection of the cyber-attacks. It uses existing data to identify the same pattern between old attacks and new attacks. ML helps to identify the features and characteristics of cyber-attacks. This paper employed four different classifiers: Random Forest (RF), Decision Tree (DT), Logistic Regression (LR), and ANN (Artificial Neural Network) on the UNSW-NB15 dataset. 9-fold cross-validation is used on the dataset to determine each classifier's best accuracy. For Ann, we used four layers which are the input layer, two hidden layers, and the output layer. In the input layer, we used 10 neurons. In the first hidden layer, we used 15 neurons, and 10 neurons for second layer. 1 neuron is used in the output layer. The classifiers are evaluated on the basis of accuracy, precision, recall, fl-score, mean square error(MSE), true positive rate (TPR), and false positive rate (FPR). Among all of the classifiers, RF overpowered other models and gave an accuracy of 0.95167% with a TPR of 0.96252% and a FPR of 0.06749%.