“School of Cognitive”
Back to Papers HomeBack to Papers of School of Cognitive
| Paper IPM / Cognitive / 11399 |
|
||||||
| Abstract: | |||||||
|
In this paper, a framework for incremental learning of belief functions, as understood in Smets? TBM, is introduced. The presented structure is discussed as general formulation of the revision mechanisms in the TBM framework, also in different levels of approximation, as the machinery for updating belief structures in online learning systems. Therefore, similar to the basic simple methods of revision (also more specific such as discounting and de-discounting operators), it is justified based on the generalized Bayesian theorem. Generality of the method and its convergence in different special cases through defining different reference functions based on the trajectories imposed by them are analyzed and ultimately, some applications are examined.
Download TeX format |
|||||||
| back to top | |||||||
