Inductive learning from incomplete and imprecise examples
We present an application of fuzzy logic with linguistic quantifiers, mainly of its calculus of linguistically quantified propositions due to Zadeh, in inductive learning under imprecision and errors. The classification into the positive and negative examples is allowed to be to a degrees (of positiveness and negativeness), between 0 and 1. The value of an attribute in an object and in a selector need not be the same allowing for an inexact matching. Errors in the data may exist though their number may be not precisely known. A new inductive learning problem is formulated as to find a concept description which best satisfies, say, almost all of the positive examples and almost none of the negative ones.
Keywordsmachine learning inductive learning learning from examples imprecision linguistic quantifier fuzzy logic
Unable to display preview. Download preview PDF.
- Bergadano F. and R. Bisio (1988) Constructive learning with continous — valued attributes. In B. Bouchon, L. Saitta and R.R. Yager (Eds.): Uncertainty and Intelligent Systems, Springer — Verlag, Berlin — Heidelberg — New York, 154–162.Google Scholar
- Bouchon B., L. Saitta and R.R. Yager (1988) Uncertainty and Intelligent Systems [Proceedings of the Second International Conference on Information Processing and Management of Uncertainty in Knowledge — Based Systems (IPMU' 88); Urbino, Italy, 1988], Springer — Verlag, Lecture Notes in Computer Science, Berlin — Heidelberg — New York.Google Scholar
- Cohen P.R. and E.A. Feigenbaum (1982) The Handbook of Artificial Intelligence. Vol. 3. Kaufmann, Los Altos.Google Scholar
- Dietterich T.G. et al. (1981) Learning and inductive inference. In P.R. Cohen and E.A. Feigenbaum (Eds.): The Handbook of Artificial Intelligence, Kaufmann, Los Altos, 323–525.Google Scholar
- Gemello R. and F. Mana (1988) Controlling inductive search in RIGEL learning system. In B. Bouchon, L. Saitta and R.R. Yager (Eds.): Uncertainty and Intelligent Systems, Springer — Verlag, Berlin — Heidelberg — New York, 171–178.Google Scholar
- Kacprzyk J. (1987) Towards ‘human — consistent’ decision support systems through commonsense — knowledge — based decision making and control models: a fuzzy logic approach. Computers and Artificial Intelligence 6, 97–122.Google Scholar
- Michalski R.S. (1973) Discovering classification rules using variable — valued logic system VL1. Proc. of the 3rd Int. Joint Conference on Artificial Intelligence (IJCAI), 162–172.Google Scholar
- Michalski R.S. (1980) Pattern recognition as rule — guided inductive inference. IEEE Trans. on Pattern Analysis and Machine Intelligence PAMI-2, 249–361.Google Scholar
- Michalski R.S. (1983) A theory and methodology of inductive learning. In R.S. Michalski, J. Carbonell and T. Mitchell (Eds.): Machine Learning. Tioga Press, Palo Alto, 83–133.Google Scholar
- Quinlan J.R. (1983) Learning efficient classification procedures and their applications to chess and games. In R.S. Michalski, J. Carbonell and T.M. Mitchell (Eds.): Machine Learning. Tioga Press, Palo Alto, 463–482.Google Scholar
- Raś Z.W. and M. Zemankova (1988) Learning driven by the concepts structure. In B. Bouchon, L. Saitta and R.R. Yager (Eds.): Uncertainty and Intellegent Systems, Springer — Verlag, Berlin — Heidelberg — New York, 193–200.Google Scholar