Robustness Analysis of Bayesian Networks with Finitely Generated Convex Sets of Distributions

Fabio Cozman

A postscript version of this paper is available.

This work appeared as technical report CMU-RI-TR-96-41, Robotics Institute, Carnegie Mellon University.

• Introduction
• Motivation
• Background
• Quasi-Bayesian theory
• Finitely generated convex sets of distributions
• Exact algorithms and the basic transformation
• Complexity and generic approximations
• Approximate inferences through parameter estimation
• Approximate inferences through Lavine's bracketing algorithm
• Classes of finitely generated convex sets of distributions
• Expected utility and variance
• Conclusion
• References