Netman: a learning network traffic controller

ABSTRACT

One of the goals of Machine Learning is the production of software that can improve itself. Such software can learn from experience and adapt to changing situations and requirements. In addition, such software can refine its knowledge-base, perhaps leading to a level of expertise beyond that of human experts. This paper describes NETMAN, a knowledge-based program that uses a machine learning technique, Knowledge-based Learning, in the domain of Network Traffic Control. NETMAN's task is to maximize call completion in a circuit-switched telecommunications network. NETMAN learns from its own experiences and by observing the actions of other agents. NETMAN is one of the components of ILS (Integrated Learning System), which contains implementations of several learning paradigms working together to improve problem-solving performance. NETMAN combines two machine learning paradigms: Explanation-Based Learning and Empirical Learning.

REFERENCES

• 1: Mark R. Adler , Alvah B Davis , Robert Weihmayer , Ralph W. Worrest, Conflict resolution strategies for nonhierarchical distributed agents, Distributed artificial intelligence: vol. 2, Morgan Kaufmann Publishers Inc., San Francisco, CA, 1990
• 2: Brandau, R. and Weihmayer, R. Heterogeneous Multiagent Cooperative Problem Solving in a Telecommunication Network Management Domain. In Benda, M. (editor), Proceedings of 9th Workshop on Distributed Artificial Intelligence, pages 41-58. American Association for Artificial Intelligence, 1989.
• 3: Alan Bundy , Bernard Silver , Dave Plummer, An analytical comparison of some rule-learning programs, Artificial Intelligence, v.27 n.2, p.137-181, Nov. 1985 [doi>10.1016/0004-3702(85)90052-9]
• 4: Danyluk, A.P. The Use of Explanations for Similarity- Based Learning. In McDermott, J. (editor), Proceedings of the Tenth IJCAI, pages 274-276. International Joint Conference on Artificial Intelligence, 1987.
• 5: Gerald Dejong , Raymond Mooney, Explanation-Based Learning: An Alternative View, Machine Learning, v.1 n.2, p.145-176, 1986 [doi>10.1023/A:1022898111663]
• 6: Pat Langley, Unifying themes in empirical and explanation-based learning, Proceedings of the sixth international workshop on Machine learning, p.2-4, December 1989, Ithaca, New York, United States
• 7: Frawley, W.J. Using Functions to Encode Domain and Contextual Knowledge in Statistical Induction. In Piatetsky- Shapiro, G. and Frawlcy, WJ. (editors), Knowledge Discovery in Databases (IJCAI-89 Workshop), pages 99-108. international Joint Conference on Artificial Intelligence, 1989.
• 8: Frawley, W.J., Fawcett, T.E. and Bradford, K. NETSIM: An object, oriented simulation of the operation and control of a circuit-switched ne~,ork. Technical Report TN 88-506. I, Computer and Intelligent Systems Laboratory, GTE Laboratories Incorporated, 1988.
• 9: Glenn A. Iba, A Heuristic Approach to the Discovery of Macro-Operators, Machine Learning, v.3 n.4, p.285-317, March 1989 [doi>10.1023/A:1022693717366]
• 10: Kopeikina, L., Bmndau, R. and Lemmon, A. Case Based Reasoning for Continous Control. In Kolodner, J (editor), Proceedings of a Workshop on Case.Based Reasoning, pages 250-259. DARPA, Morgan-Kaufmann, 1988.
• 11: Perry M. Kosieniak , Victor Mathis , Michael St. Jacques , Delano Stevens, The network control assistant (NCA), a real-time prototype expert system for network management, Proceedings of the 1st international conference on Industrial and engineering applications of artificial intelligence and expert systems, p.367-379, June 1988, Tullahoma, Tennessee, United States [doi>10.1145/51909.51951]
• 12: Michael Lebowitz, Experiments with Incremental Concept Formation: UNIMEM, Machine Learning, v.2 n.2, p.103-138, September 1987 [doi>10.1023/A:1022800624210]
• 13: Tom M. Mitchell , Richard M. Keller , Smadar T. Kedar-Cabelli, Explanation-Based Generalization: A Unifying View, Machine Learning, v.1 n.1, p.47-80 [doi>10.1023/A:1022691120807]
• 14: Silver, B. Precondition Analysis: Learning Control Informarion. In Michalski, R.S., Carbonell, J.G. and Mitchell, T.M. (editors), Machine Learning: An Artificial intelligence Approach Vol 2, pages 647-670. Morgan-Kaufmann, 1986.
• 15: Silver, B. Studies in Computer Science and Artificial Intelligence. Number 1: Meta-Level Inference. North Holland, 1986
• 16: Silver, B. A Hybrid Approach in an Imperfect Domain. In i)eJong, G (editor), Proceedings of AAAI Symposium on Explanation-Based Learning. American Association for Arfifici~ Intelligence, 1988.
• 17: Bernard Silver , William Frawley , Glenn Iba , John Vittal , Kelly Bradford, ILS: a framework for multi-paradigmatic learning, Proceedings of the seventh international conference (1990) on Machine learning, p.348-356, June 1990, Austin, Texas, United States
• 18: Bernard Silver , John Vittal , William Frawley , Glenn Iba , Tom Fawcett , Susan Dusseault , John Doleac, A framework for integrating heterogeneous learning agents, Second generation expert systems, Springer-Verlag New York, Inc., Secaucus, NJ, 1993

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.6 Learning
          Subjects: Knowledge acquisition

Additional Classification:
  C. Computer Systems Organization
  C.2 COMPUTER-COMMUNICATION NETWORKS
      C.2.1 Network Architecture and Design
          Subjects: Circuit-switching networks

  F. Theory of Computation
  F.2 ANALYSIS OF ALGORITHMS AND PROBLEM COMPLEXITY
      F.2.2 Nonnumerical Algorithms and Problems
          Subjects: Routing and layout

  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.2 Automatic Programming
          Subjects: Program transformation

General Terms:
Algorithms, Design