Proposal of Online Outlier Detection in Sensor Data Using Kernel Density Estimation

Sensors in different locations can generate streaming data, which can be analyzed in real-time to identify events of interest. Continuous outlier detection in data streams has important applications in fraud detection, network security, environmental monitoring and public health. In this paper, we propose a framework that computes in a distributed manner an approximation of multi-dimensional data distributions in order to enable complex applications in resource constrained sensor networks. Here we are targeting the problem of outlier detection. We demonstrate how our technique can be used to identify either distance based or density based outliers in a single pass over the data. Our approach takes into consideration various characteristics and features of streaming sensor data.