A Novel Framework for Discovering Anomaly Detection Errors in IoT at Real Time

Internet of Things (IoT) is a new paradigm that recently gained popularity. IoT is still in its infancy, so it faces a lot of obstacles varying from security, privacy and data management issues. One main challenge is related to anomaly detection. An anomaly is a data value which is far away from the remaining data values. Discovering anomaly detection errors in IoT in real time is an existing problem that needs to be explored. Most of the anomaly detection algorithms in IoT suffer from detection errors such false positives and false negatives. These detection errors affect the efficiency of the anomaly detection techniques in correctly detecting the anomalies and hence constrain IoT applications in taking the correct actions at a suitable time. As a result, an approach is needed to discover the false positives and false negatives at real time so that the necessary actions should be taken quickly. Consequently, in this paper, a framework is proposed to discover the anomaly detection errors in IoT in real time. This framework can be applied on any exiting anomaly detection algorithm to detect the false positives and false negatives in real time.