| Popis: |
Among the many challenges in the Big Data Security field is that we have to secure a high volume and velocity flow of data, this cannot be done using classic static security methods, we need a different approach, one that can analyze the traffic in real time and gives fast feedback and one of these approaches is Anomaly detection. Anomaly detection or outlier detection has been an active research area for the past years, it is an automatic and reactive method that can be used to flag any suspicious behavior in the traffic of Big Data systems. In this paper we present a benchmark study of the most used Anomaly detection algorithms in the literature. We focus on analyzing their performance on detecting anomalies in Big Data. The performance of the algorithms will be evaluated on the UNSW-NB15 Dataset. This study aims to determine what is the best algorithm or group of algorithms that performs very well in a Big Data context. |
