Deep Learning Model for Intrusion Identification

Mohasin B. Tamboli and Dr. Nageswara Rao Moparthi

In today’s era, Intrusion detection system (IDS) is playing major role in cyber security to protecting computer resources from external attacks. Unpredictable behavior and an unknown vulnerability to advanced attacks have created a major cyber security challenge. Various IDS models have been developed with datasets such as KDD, KDD-CPU99, and KDD upgraded versions. This paper discus an Deep Learning model (DLM) to detect network attacks using CICIDS-2017 dataset. The performance of DLM has been compared with different classification algorithms, such as Random Forecast (RF) classifier, Support Vector Machine, and Naïve bias models. Experiment results shows that DLM model has higher performance over other models when evaluating data with CICIDS-2017 dataset.

Volume 12 | Issue 5

Pages: 388-395

DOI: 10.5373/JARDCS/V12I5/20201726