NAAD using cross validation

Network Activity Anomaly Detection : I have trained and evaluated four models: Logistic Regression, Random Forest, SVM, Neural Network using cross-validation and F1 scores. The neural network's training and validation losses were tracked separately. I compared model performances through ROC curves, AUC scores, and confusion matrices. As a conclusion, i find that all the models are performing outstanding but logistic regression suffered to predict only 2 entries than others. Submitting the output of random forest.