Supervised UFR (UFR Fast Regression) Machine Learning Algorithm for Enhancing Performance of Intrusion Detection System
J. Dillibabu1, K. Nirmala2
1J. Dillibabu, Research scholar Bharathiyar University, Coimbatore,Tamilnadu, India.
2K. Nirmala, Associate Professor, Quaid-E-Millath Government college for women, Chennai, India.
Manuscript received on 02 June 2019 | Revised Manuscript received on 10 June 2019 | Manuscript published on 30 June 2019 | PP: 943-949 | Volume-8 Issue-8, June 2019 | Retrieval Number: G5652058719/19©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: Cloud platform and data centers rely on forecasting for predicting future workload accuracy. Forecasting information facilitate appropriate virtualization in provisioning infrastructure in cost-effective manner. Forecasting accuracy relies on underlying algorithm performance merit based on data fed in the network. Data analyst faces problem of timely identification of discontinuities in data forecasting. Discontinuity is referred as abrupt time-series pattern change for performance persists which cannot be recur. Data analyst must identify the non-sequences in the data performance before proposing forecast method and retain forecasting model data performance at the time of data synchronization. In existing numerous approaches and tools are available for anomaly detection in data. However, still automated tools for anomaly detection do not exist for anomaly detection in data centre. This paper introduces a unsupervised fast regression (UFR) model for discontinuity data detection in large and small data centers. UFR model combines concept of regression where data are converted in to binary form for data discontinuity detection. The proposed approach facilitates cloud providers and data center analysts automatically detect data discontinuity. Performance evaluation of proposed approach with existing approach exhibited significant performance in terms of discontinuities identification with increased accuracy.
Keyword: Unsupervised fast regression (UFR), Malicious Activity, Security, Discontinuity.
Scope of the Article: Machine Learning.