Using Bayesian Networks to evaluate the trustworthiness of ‘2 out of 3’ decision fusion mechanisms in multi-sensor applications

Abstract The use of smart-sensors to recognize automatically complex situations (anomalous behaviors, physical security threats, etc.) requires ‘intelligent’ methods to improve the trustworthiness of automatic decisions. Voting and consensus mechanisms can be employed whether supported by probabilistic formalisms to correlate event occurrence, to merge local events and to estimate the likelihood of overall decisions. This paper presents the results of a quantitative comparison of three different voting schemes based on Bayesian Networks. These models present a growing complexity and they are able to provide a trustworthiness estimation based on single nodes detection reliability in terms of false alarm probabilities.

[1]  Pramod K. Varshney,et al.  Distributed detection with multiple sensors I. Fundamentals , 1997, Proc. IEEE.

[2]  R. Polikar,et al.  Ensemble based systems in decision making , 2006, IEEE Circuits and Systems Magazine.

[3]  Francesco Flammini,et al.  Dependable integrated surveillance systems for the physical security of metro railways , 2009, 2009 Third ACM/IEEE International Conference on Distributed Smart Cameras (ICDSC).

[4]  Tom Fawcett,et al.  An introduction to ROC analysis , 2006, Pattern Recognit. Lett..

[5]  Ole J. Mengshoel,et al.  Sensor Validation using Bayesian Networks , 2008 .

[6]  Bogdan Gabrys,et al.  Classifier selection for majority voting , 2005, Inf. Fusion.

[7]  Klamer Schutte,et al.  A comparison of decision-level sensor-fusion methods for anti-personnel landmine detection , 2001, Inf. Fusion.

[8]  Stefano Marrone,et al.  A new modeling approach to the safety evaluation of N-modular redundant computer systems in presence of imperfect maintenance , 2009, Reliab. Eng. Syst. Saf..

[9]  Francesco Flammini,et al.  DETECT: a novel framework for the detection of attacks to critical infrastructures , 2008 .

[10]  Behrooz Parhami,et al.  A taxonomy of voting schemes for data fusion and dependable computation , 1996 .

[11]  Stefano Marrone,et al.  Trustworthiness Evaluation of Multi-sensor Situation Recognition in Transit Surveillance Scenarios , 2013, CD-ARES Workshops.

[12]  Luigi Portinale,et al.  Bayesian networks in reliability , 2007, Reliab. Eng. Syst. Saf..

[13]  Kevin Murphy,et al.  Bayes net toolbox for Matlab , 1999 .