An Experimental Analysis and Performance Evaluation of Machine Learning Classifiers using Benchmark Dataset
Sukhvinder Singh1, S K V Jayakumar2
1Sukhvinder Singh, School of Engineering and Technology, Department of Computer Science, Pondicherry University, Puducherry, India.
2S K V Jayakumar, School of Engineering and Technology, Department of Computer Science, Pondicherry University, Puducherry, India.
Manuscript received on 13 April 2019 | Revised Manuscript received on 18 May 2019 | Manuscript published on 30 May 2019 | PP: 2425-2436 | Volume-8 Issue-1, May 2019 | Retrieval Number: A1982058119/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© 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: In the world of system automation, organizations are really struggling to prevent/identify the different types of attacks or suspicious activities happening in the network because interconnected networks are widely spread across the globe. The organizations must have a virtuous traffic analysis system for fast identification of malicious traffic and isolating the attacker at the early stage, which demands an experimentally proven and analyzed robust traffic classification technique. There are many methods available to classify the traffic such as Port Number Based, Flow-Based and Packet-Based. Machine Learning is one of the state-of-the-art tools to solve classification problems in the real world. As it can attain better results as compared to traditional methods. In this paper, various variants of Machine Learning techniques under Support Vector Machine (SVM), Decision Tree and Ensemble is discussed. Further, Principal Component Analysis (PCA) is used for the dataset feature extraction and entrenched into the classifiers for traffic classification. The experimental analysis illustrates the highest accuracy rate of 99.7% for traffic classification by Bagged Tree classifier one of the variants of Ensemble as compared to other variants of SVM, Decision Tree, and Ensemble Classifiers.
Index Terms: Decision Tree, Denial of Service, Ensemble, Machine Learning, Principal Component Analysis, Support Vector Machine.
Scope of the Article: Machine Learning