Network Intrusion Detection using Machine Learning Techniques
I. Sumaiya Thaseen, Babu Poorva, Pamidi Sai Ushasree
Abstract
Intrusion detection over packets on a network aims at classifying different types of packets without decrypting its contents to detect any intrusions in the network using machine learning. For this, packets are generated and transmitted over a network which are then captured by Wireshark for intrusion detection analysis. The captured data is organized into a dataset with the selected attributes after preprocessing using Weka tool and machine learning algorithms such as Naive Bayes, Support vector machine, Random Forest and KNearestNeighbors are implemented which classifies the data with accuracy 83.63%, 98.23%, 99.81%, 95.13% respectively. The packets are classified as encrypted packets, unencrypted packets, unencrypted malicious packets and encrypted malicious packets with different attributes and features where random forest is identified to be the best classifier with maximum accuracy.