A fast and efficient anomaly detection is a major recent research challenge due to the increasing usage of internet based services. Anomaly detection can effectively help in finding the fraud, finding abnormal action in huge and complex Big Data sets. The paper concentrate on anomaly detection and introduced a novel machine learning based anomaly detection in big data. Machine learning for anomaly detection helps to detect anomalies from large data sets and enhance the speed of detection. We evaluate the proposed framework with large data sets by performing experiments and compare with several existing anomaly detection techniques. The proposed models gives experimental results in terms of detection rate (96.86%), false positive rate (1.297%), accuracy (98.72%) and F-Measure (98.30%).
Volume 11 | 06-Special Issue
Pages: 518-524