An Enhanced Intelligent Intrusion Detection System using Machine Learning
Dhikhi T1, M.S. Saravanan2
1Dhikhi, Research Scholar in Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, Chennai. She is currently Assistant Professor with SRM Institute of Science and Technology, Ramapuram, Chennai India.
2Dr. M.S. Saravanan, Associate Professor in the Department of Computer Science and Engineering, Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, Chennai, India.
Manuscript received on 27 June 2019 | Revised Manuscript received on 05 July 2019 | Manuscript published on 30 July 2019 | PP: 2177-2181 | Volume-8 Issue-9, July 2019 | Retrieval Number: H6932068819/19©BEIESP | DOI: 10.35940/ijitee.H6932.078919
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: Numerous Intrusion detection techniques are used to find the anomalies that depends on the accuracy, detection rate etc. The purpose of the system is to detect the anomalies based on the given dataset thereby improving the accuracy. A CWS IDS is proposed to find the anomalies in the network, that combines machine learning techniques autoencoder and support vector machine for feature extraction and classification. This is evaluated on the training and testing datasets of NSL KDD dataset that accomplishes well in terms of reduction rate and precision. By combining autoencoder and support vector machine for finding the anomalies, the performance metrics of the system is improved.The system is related with single SVM and Random forest classifier. The performance measures such as precision, recall, accuracy and F-measure is equated with the SVM, random forest, and CWS IDS for training data and test data. Thereby the recognition rate is enhanced and both false positives, false negatives are lesser.
Keywords: Contractive Encoder, Intrusion Detection, NSL-KDD Dataset, Support Vector Machin.
Scope of the Article: Machine Learning