TSNet – A Distributed Architecture for Time Series Analysis

Abstracting complex time-series data is a necessary precursor to the application of clinical guidelines and protocols in areas such as the Intensive Care Unit where large volumes of data are generated. If we are to obtain a consensus as to the best ways to achieve these abstractions, different research groups need to be able to experiment with data acquired from a variety of sources, and to apply algorithms developed elsewhere. However, sharing of data and algorithms is known to be difficult. This paper sets out an infrastructure which has been developed to allow collaborative working of this kind.