Optimization and Stream Data Processing

Janusz Granat^*

Event-based computing is becoming a central aspect of emerging large-scale distributed computing paradigms. One of the components of such systems is set of algorithms for detection of events based on data streams. There are various research areas related detection of events like change point detection, anomaly detection or novelty detection. We will present application of stochastic optimization for event detection in data stream. The essential challenge of the algorithm development was taking into account the differences between data stream processing and traditional data processing in particular the limitation of number of passes of data processing, processing time and memory usage. The applications from the field of healthcare (on-line patient monitoring) and telecommunication will be presented.

Mathematics Subject Classification: 90X08

Keywords: Event mining, complex events processing, decision support

Minisymposion: Integration of Optimization, Modeling and Data Analysis for Solving Real World Problems