Comparison of Bayesian networks and Dempster-Shafer theory in attribute tracking systems

Thesis Type: Postgraduate

Institution Of The Thesis: Orta Doğu Teknik Üniversitesi, Faculty of Engineering, Department of Electrical and Electronics Engineering, Turkey

Approval Date: 2013




In modern civil and military avionic systems, traffic control constitutes one of the most critical parts that requires high-speed, reliable and robust decisions to be done even under noisy conditions. In conjunction with the rapidly developing computer technology, systems became available to perform very long processes within milliseconds. In this thesis, probabilistic models will be used in order to classify a target and detect if the target is making use of electronic counter-measures (ECM). Thereafter, the performances of different systems under the same conditions will be compared. Bayesian Networks Theory and Dempster-Shafer Evidence Theory are two most well-known and applicable approaches to classification and attribute tracking problems. Therefore, aforementioned two approaches are chosen in order to simulate desired attribute tracking and detection scenarios. Subsequent to presenting results obtained by applying abovementioned theories to the selected scenarios, improvements are made in order to increase system performance. The effects of quality of the information source and improvements are presented within this thesis as well as a general comparison of implemented theories.