This paper reviews existing forms of density-based, partitional and hierarchical clustering methods in the context of flight data analysis. Advantages and disadvantages are fully explored with a focus on proposing a clustering-based ensemble framework for monitoring flight data in order to search for anomalies during flight operation. Case studies in selected flight scenarios are provided to demonstrate the potential of clustering methods and their integration with reasoning techniques in detecting abnormal flights.
|Title of host publication||Knowledge-based intelligent information and engineering systems: 12th International Conference, KES 2008, Zagreb, Croatia, September 3-5, 2008, proceedings, part I|
|Editors||I. Lovrek, R. Howlett, L. Jain|
|Place of Publication||Berlin|
|Number of pages||8|
|Publication status||Published - 2008|
|Name||Lecture notes in artificial intelligence|