Performance and Computation Time Enhancement of Various Machine Learning Techniques for NSL-KDD Dataset
Pradeep K V1, K Anusha2, S. Nachiyappan3
1Pradeep K V*, Asst. Prof, SCOPE, VIT University, Chennai.
2Anusha K. Assoc. Prof. SCOPE, VIT University, Chennai.
3Nachiyappan S. Asst. Prof.(Sr), SCOPE, VIT University, Chennai.
Manuscript received on March 15, 2020. | Revised Manuscript received on March 24, 2020. | Manuscript published on March 30, 2020. | PP: 4726-4730 | Volume-8 Issue-6, March 2020. | Retrieval Number: F9877038620/2020©BEIESP | DOI: 10.35940/ijrte.F9877.038620
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: To develop an effective intrusion detection system we definitely need a standardize dataset with a huge number of correct instances without missing values. This would significantly help the system to train and test for real-time use. Previously for research purpose, KDD-CUP’99 dataset has been used, but later on, it has been seen that it is not so useful for training the model as it consists a lot of missing and abundant values. All this issue have been tackled in NSL dataset. To analyze the capabilities of the dataset for intrusion detection system we have analyzed various machine learning classification algorithm to classify the attack over any network. This paper has explored many facts about the dataset and the computation time.
Keywords: KDD, Computation Time, Intrusion Detection.
Scope of the Article: Machine Learning.