Automated network troubleshooting knowledge acquisition

Abstract

Troubleshooting knowledge acquisition is a notorious network maintenance expert systems development bottleneck. We present an improved methodology to generate automatically a skeleton of network troubleshooting knowledge base given the data about network topology, test costs, and network component failure likelihood. Our methodology uses AO* search where a suitable modification of the Huffman code procedure is found to be an admissible heuristic. Our heuristic uses synergistically information about both component failure rates and test costs while relaxing topology constraints. The resulting expert system (XTAR) minimizes expected troubleshooting cost faster and learns better troubleshooting techniques during its operation.

Cost-Sensitive Classification: Empirical Evaluation of a Hybrid Genetic Decision Tree Induction Algorithm

“…However, searching for the optimal solution is computationally infeasible (Pearl, 1988). We attempted to take a decision theoretic approach to this problem by implementing the AO* algorithm (Pearl, 1984) and designing a heuristic evaluation function to speed up the AO* search (Lirov & Yue, 1991). We were unable to make this approach execute fast enough to be practical.…”

“…The problem of cost-sensitive classification arises frequently. It is a problem in medical diagnosis (Núñez, 1988(Núñez, , 1991, robotics (Tan & Schlimmer, 1989Tan, 1993), industrial production processes (Verdenius, 1991), communication network troubleshooting (Lirov & Yue, 1991), machinery diagnosis (where the main cost is skilled labor), automated testing of electronic equipment (where the main cost is time), and many other areas.…”

Cost‐sensitive methods of constructing hierarchical classifiers

“…One of the main tasks of designing computeraided diagnosis is finding the balance between cost of exploitation and qualities of diagnosis. As we stated the problem of cost-sensitive decision making arises frequently in medicine (Nu´nez, 1988), industrial production processes (Verdenius, 1991), robotics (Tan & Schlimmer, 1989), technological diagnosis (Lirov & Yue, 1991) and many other fields, e.g. electronic equipment testing and real-time computer systems.…”