Entropy and information theory
Entropy and information theory
Decision-theoretic troubleshooting
Communications of the ACM
Introduction to Bayesian Networks
Introduction to Bayesian Networks
Heuristics for Two Extensions of Basic Troubleshooting
SCAI '01 Proceedings of the Seventh Scandinavian Conference on Artificial Intelligence
Artificial Intelligence: A Modern Approach
Artificial Intelligence: A Modern Approach
Automated Planning: Theory & Practice
Automated Planning: Theory & Practice
AND/OR search spaces for graphical models
Artificial Intelligence
Decision-theoretic troubleshooting: a framework for repair and experiment
UAI'96 Proceedings of the Twelfth international conference on Uncertainty in artificial intelligence
When to test? Troubleshooting with postponed system test
Expert Systems with Applications: An International Journal
| Hi-index | 0.01 |
We propose a troubleshooting algorithm that can troubleshoot systems with dependent action costs. When actions are performed they may change the way the system is decomposed and affect the cost of future actions. We present a way to model this by extending the traditional troubleshooting model with an additional state that describes which parts of the system that are decomposed. The proposed troubleshooting algorithm searches an AND/OR graph with the aim of finding the repair plan that minimizes the expected cost of repair. We present the heuristics needed to speed up the search and make it competitive with other troubleshooting algorithms. Finally, the performance of the algorithm is evaluated on a probabilistic model of a fuel injection system of a truck. We show that the expected cost of repair can be reduced when compared with an algorithm from previous literature.