Dynamic update of data analysis models in emergency systems

The analysis of sensor data is a major activity in an emergency system; it is aimed at extracting useful information and at executing monitoring and anomaly detection. We focus on automatic data analysis through machine learning techniques, which require creating a model of the data that has to be kept up to date to match the evolving status of the environment. The update of a model improves its quality but introduces computation and communication overhead. In this paper we address the problem of identifying the optimal trade off between a low update rate and high quality of the model, we describe two update strategies and we draw considerations from their application on two sets of sensor data.