“School of Cognitive”

Back to Papers Home
Back to Papers of School of Cognitive

Paper   IPM / Cognitive / 11398
School of Cognitive Sciences
  Title:   Modeling Sensor Reliabilities in the Belief Functions Framework for Learning by Evolutionary Algorithms
  Author(s): 
1.  Taher Shahbazi
2.  Babak Nadjar Arabi
3.  Majid Nili Ahmadabadi
4.  Caro Lucas
  Status:   In Proceedings
  Proceeding: 1st Joint Congress on Fuzzy and Intelligent Systems, Mashhad, Iran, 2007
  Year:  2007
  Supported by:  IPM
  Abstract:
Modeling and assessing sensor reliabilities is one of the most important problems in the information fusion. Learning reliability coefficients in the uncertainty model of a distributed sensor network is a difficult nonlinear optimization problem. However, utilizing more flexible models such as belief functions and considering precise error functions leads to achieve more accurate solutions, makes the problem more complicated. Evolutionary algorithms are known to be powerful tools to solve nonlinear optimization problems. Above mentioned facts about the learning of sensor reliabilities make the evolutionary algorithms a suitable option. Population Based Incremental Learning (PBIL) is a rather new evolutionary algorithm which works better than genetic algorithms in most of the optimization problems. In this paper, we develop an extension of the standard PBIL algorithm to continuous search spaces for the constrained nonlinear optimization problem of the learning sensor reliabilities. We apply this algorithm to decision fusion based on utility function and examine its performance over a synthetic data.

Download TeX format
back to top
Clients Logo
Clients Logo
Clients Logo
Clients Logo
Clients Logo
Clients Logo
Clients Logo
Clients Logo
scroll left or right