2012 | OriginalPaper | Buchkapitel
Evolving Event Detectors in Multi-channel Sensor Data
verfasst von : Feng Xie, Andy Song, Vic Ciesielski
Erschienen in: AI 2012: Advances in Artificial Intelligence
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Detecting events of interest in sensor data is crucial in many areas such as medical monitoring by body sensors. Current methods often require prior domain knowledge to be available. Moreover, it is difficult for them to find complex temporal patterns existing in multi-channel data. To overcome these drawbacks, we propose a Genetic Programming (GP) based event detection methodology which can directly take raw multi-channel data as input. By applying it to three event detection tasks with various event sizes and comparing with four typical classification methods, we can see that those detectors evolved by GP can handle raw data much better than other methods. With features manually defined based on domain knowledge, our method can also be comparable with others. The analysis of evolved detectors demonstrates that distinctive characteristics of the target events are captured by these GP detectors.