Abstract
This paper shows, by discussing a number of Machine Learning (ML) applications, that the existing ML techniques can be effectively applied in knowledge acquisition for expert systems, thereby alleviating the known knowledge acquisition bottleneck. Analysis in domains of practical interest indicates that the performance accuracy of knowledge induced through learning from examples compares very favourably with the accuracy of best human experts. Also, in addition to accuracy, there are encouraging examples regarding the clarity and meaningfulness of induced knowledge. This points towards automated knowledge synthesis, although much further research is needed in this direction. The state of the art of some approaches to Machine Learning is assessed relative to their practical applicability and the characteristics of a problem domain.
Similar content being viewed by others
References
Bratko, I. and Kononenko, I., “Learning Diagnostic Rules from Incomplete and Noisy Data,” inInteractions in AI and Statistics (B. Phelps, ed,), Gower Technical Press, London, 1987.
Bratko, I., Mozetič I. and Lavrač, N.,KARDIO: A Study in Deep and Qualitative Knowledge for Expert Systems, The MIT Press, Cambridge, Massachusetts, 1989.
Bratko, I., Muggleton, S. and Varšek, A., “Learning Qualitative Models of Dynamic Systems,”Proc. Inductive Logic Programming ILP-91, Viana do Castelo, Portugal, March 1991. (Also in Ref. 34); abbreviated version inMachine Learning: Proc. Eighth Int. Workshop (L. A. Birnbaum and G. Collins, eds.), Morgan Kaufmann, CA).
Bratko, I. and Mulec, P., “An Experiment in Automatic Learning of Diagnostic Rules,”Informatica, 4, 4, pp. 353–359, 1979.
Cestnik, B. and Bratko, I., “Learning Redundant Rules in Noisy Domains,”Proc. ECAI 88, European Conf. on Artificial Intelligence, Muenchen, August 1988.
Breiman, L., Friedman, J. H., Olshen, R. A. and Stone, C. J.,Classification and Regression Trees, Wadsworth, Belmont, CA, 1984.
Cestnik, B. and Bratko, I., “On Estimating Probability in Decision Tree Pruning,”Proc. EWSL-91, Porto, Portugal, March 1991, Springer-Verlag, 1991.
Cestnik, B., Kononenko, I. and Bratko, I., “ASSISTANT 86: A Knowledge Elicitation Tool for Sophisticated Users,” inProgress in Machine Learning (I. Bratko, N. Lavrač, eds.;Proc. Second European Working Session on Learning, Bled, Slovenia, 1987), Sigma Press, Wilmslow, England, 1987.
Chandrasekaran, B., “Interview with Frederick Hayes-Roth and Richards Fikes,”IEEE Expert, pp. 3–14, October 1991.
Chilausky, R., Jacobsen, B. and Michalski, R. S., “An Application of Variable-Valued Logic to Inductive Learning of Plant Disease Diagnostic Rules,”Proceedings of the 6th International Symposium on Multiple-Valued Logic, May, 25–28, 1976, Logan, Utah, 1976.
Clark, P., Cestnik, B., Sammut, C. and Stender, J., “Applications of Machine Learning: Notes from the Panel Members,”Machine Learning — EWSL-91; Proc. European Working Session on Learning, Porto, Portugal, March 1991, Springer-Verlag, 1991.
Clark, P. and Niblett, T., “The CN2 Induction Algorithm,”Machine Learning, 3, 4, pp. 261–284, 1989.
Coiera, E., “Generating Qualitative Models from Example Behaviours,”DCS Report, No. 8901, School of Electr. Eng. and Computer Sc., Univ. of New South Wales, Sydney, Australia, 1989.
Dolšak, B. and Muggleton, S. “The Application of Inductive Logic Programming to Finite Element Mesh Design,”Proc. ILP-91, Viana do Castelo, Portugal (also in Ref. 34)), 1991.
Dolšak, B., “Determining the Geometric Model of Finite Element Meshes Using AI Methods,”M. Sc. thesis, University of Maribor: CAD Center, Maribor, Slovenia, 1991.
Gams, M., “New Measurements Highlight the Importance of Redundant Knowledge,”Proc. 4th. European Working Session on Learning (K. Morik, ed.; Montpellier, France), Pitman and Morgan Kaufmann, 1989.
Goldman, M. J.,Principles of Clinical Electrocardiography, Lange Medical Publications, Los Altos, 1976.
Kononenko, I. and Bratko, I., “Information Based Evaluation Criterion for Classifier’s Performance,”Machine Learning, 6, pp. 67–80, 1991.
Krann, I., Richards, B. and Kuipers, B., “Automatic Abduction of Qualitative Models,”Proc. QR-91) Austin, Texas, May 1991.
Lavrač, N., Džeroski, S. and Grobelnik, M., “Learning Nonrecursive Definitions of Relations with LINUS,”EWSL ’91: Machine Learning: Proceedings of the European Working Session on Learning (Y. Kodratoff, ed.), Porto, Portugal,Lecture Notes in Artificial Intelligence, Springer-Verlag, 1991.
Leech, W. J., “A Rule Based Process Control Method with Feedback,” inAdvances in Instrumentation, 41, Part 1, 69–175, 1986.
Michalski, R. S., “A Theory and Methodology of Inductive learning,” inMachine Learning: An Artificial Intelligence Approach (R.S. Michalski, J. G. Carbonell and T. M. Mitchell, eds.), Tioga, Palo Alto, CA, 1983.
Michalski, R. S. and Chilausky, R. L., “Learning by Being Told and Learning from Examples: An experimental Comparison of the Two Methods of Knowledge Acquisition in the Context of Developing an Expert System for Soybean Disease Diagnosis,”International Journal of Policy Analysis and Information Systems, 4, pp. 125–161, 1980.
Michie, D., “The Superarticulacy Phenomenon in the Context of Software Manufacture,”Proceedings of the Royal Society A, 405, London, pp. 185–212, 1986.
Michie, D., “Machine Learning in the Next Five Years,”EWSL-88-Proc. 3rd European Working Session on Learning, Glasgow, 1988, Pitman, London, 1988.
Michie D., “Problems of Computer-Aided Concept Formation,” inApplications of Expert Systems, 2 (J. R. Quinlan, ed.), Turing Institute Press in association with Addison-Wesley, 1989.
Michie, D., “Personal Models of Rationality,”Journal of Statistical Planning and Inference, 25, pp. 381–399, 1990.
Michie, D., Muggleton, S., Riese, C. and Zubrick, S., “RuleMaster: A Second Generation Knowledge Engineering Tool,”Proc. Ist Conf. Artificial Intelligence Applications, IEEE Computer Society, 1984.
Mingers, J., “An Empirical Comparison of Selection Measures for Decision Tree-Induction,”Machine Learning, 3, 4, pp. 319–342, 1989.
Mingers, J., “An Empiricial Comparison of Pruning Methods for Decision-Tree Induction,”Machine Learning,4,2, 1989.
Mozetič, I., “Learning of Qualitative Models,” inProgress in Machine Learning (I. Bratko, N. Lavrač, eds.Proc. 2nd European Working Session on Learning, Bled, Slovenia), Sigma Press, Wilmslow, 1987.
Mozetič, I., “The Role of Abstractions in Learning Qualitative Models,”,Proc. 4th Int. Workshop on Machine Learning, Irvine, CA, June 1987, Morgan Kaufmann, 1987.
Muggleton, S., “Inductive Logic Programming,”New Generation Computing, 8, pp. 295–318, 1991.
Muggleton, S., (ed.),Inductive Logic Programming, Academic Press, 1992.
Muggleton, S. and Buntine, W., “Machine Invention of First Order Predicates by Inverting Resolution,”Proc. 5th Int. Conf. Machine Learning, Ann Arbor, MI, 1988.
Muggleton, S. and Feng, C., “Efficient Induction of Logic Programs,”Proc. Ist Conf. Inductive Learning Theory, Tokyo, Japanese Society for Artificial Intelligence, 1990.
Muggleton, S., King, R. D. and Sternberg, M. J. E., “Protein Secondary Structure Using Logic” (to appear in Protein Eng.), 1992.
Niblett, T. and Bratko, I., “Learning Decision Trees in Noisy Domains,” inExpert Systems 86: Proc. Expert Systems 86 Conf. (M. Bramer, ed.), Cambridge Univ. Press, 1986.
Phibbs, B.,The Cardiac Arrhythmias, The C. V. Mosby. Co., St. Louis, 1973.
Pirnat, V., Kononenko, I. and Bratko, I., “Medical Analysis of Automatically Induced Diagnostic Rules,”Proc. 3rd Int. Conf. AI in Medicine, London, August 1989.
Quinlan, J. R., “Discovering Rules by Induction from Large Collections of Examples,” inExpert Systems in the Microelectronic Age (D. Michie, ed.), Edinburgh University Press, 1979.
Quinlan, J. R., “Induction of Decision Trees,”Machine Learning Journal, 1, pp. 81–106, 1986.
Quinlan, J. R., “Learning Logical Definitions from Relations,”Machine Learning, 5, pp. 239–266, 1990.
Quinlan, J.R., Compton, P., Horn, K. A. and Lazarus, L., “Inductive Knowledge Acquisition: A Case Study,” inApplications of Expert Systems, 2 (J. R. Quinlan, ed.), Turing Institute Press in association with Addison-Wesley, 1989.
Roškar, E., Abrams, P., Bratko, I., Kononenko, I. and Var<ek, A., “MCUDS —An Expert System for the Diagnostics of Lower Urinary Tract,”Journal of Biomedical Measurement Informatics and Control, 1, 4, pp. 201–204, 1986.
Slocombe, S., Moore, K. D. M. and Zelouf, M., “Engineering Expert System Applications,”Technical Report, British Petroleum, 1986.
Urbančič, T., Kononenko, I. and Križman, V., “Review of AI Applications at Ljubljana AI Labs.,”Technical Report, DP-6183, Ljubljana, J. Stefan Institute, 1991.
Author information
Authors and Affiliations
Additional information
Ivan Bratko, Ph. D.: He is professor at the Faculty of Electrical Eng. and Computer Science, Ljubljana University, Slovenia. He heads the AI laboratories at J. Stefan Institute and the University. He is also fellow of the Turing Institute, Glasgow, Scotland, and the chairman of the International School for the Synthesis of Expert Knowledge (ISSEK, based in Udine, Italy). He has conducted research in machine learning, knowledge-based systems, qualitative modelling, intelligent robotics, heuristic programming and computer chess. Currently he is the principal researcher of the project Automated Knowledge Synthesis, and is collaborating in the European Esprit III Basic Research Project Inductive Logic Programming. Ivan Bratko received his B. Sc. and Ph. D degrees from Ljubljana University, Slovenia.
About this article
Cite this article
Bratko, I. Applications of Machine Learning: Towards knowledge synthesis. New Gener Comput 11, 343–360 (1993). https://doi.org/10.1007/BF03037182
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF03037182