Generating time series reference models based on event analysis

Creating a reference model that represents a given set of time series is a relevant problem as it can be applied to a wide range of tasks like diagnosis, decision support, fraud detection, etc. In some domains, like seismography or medicine, the relevant information contained in the time series is concentrated in short periods of time called events. In this paper, we propose a technique for generating time series reference models based on the analysis of the events they contain. The proposed technique has been applied to time series from two medical domains: Electroencephalography, a neurological procedure to record the electrical activity produced by the brain and Stabilometry, a branch of medicine studying balance-related functions in human beings.