Framework for Providing Security in Private Cloud using Machine Learning Techniques
Shridhar Allagi1, Rashmi Rachh2
1Shridhar Allagi, Department of Computer Science and Engineering, KLE Institute of Technology, Hubbali, India.
2Dr. Rashmi Rachh, Department of Computer Science and Engineering, Visvesvaraya Technological University, Belagavi, India.
Manuscript received on September 23, 2019. | Revised Manuscript received on October 15, 2019. | Manuscript published on October 30, 2019. | PP: 7641-7645 | Volume-9 Issue-1, October 2019 | Retrieval Number:F9121088619/2019©BEIESP | DOI: 10.35940/ijeat.F9121.109119
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: The advancement in cyber-attack technologies have ushered in various new attacks which are difficult to detect using traditional intrusion detection systems (IDS).Existing IDS are trained to detect known patterns because of which newer attacks bypass the current IDS and go undetected. In this paper, a two level framework is proposed which can be used to detect unknown new attacks using machine learning techniques. In the first level the known types of classes for attacks are determined using supervised machine learning algorithms such as Support Vector Machine (SVM) and Neural networks (NN). The second level uses unsupervised machine learning algorithms such as K-means. The experimentation is carried out with four models with NSL- KDD dataset in Openstack cloud environment. The Model with Support Vector Machine for supervised machine learning, Gradual Feature Reduction (GFR) for feature selection and K-means for unsupervised algorithm provided the optimum efficiency of 94.56 %.
Keywords: Intrusion Detection System (IDS), Support Vector Machine (SVM), Supervised Machine Learning, Unsupervised Machine Learning, Gradual Feature Reduction (GFR).