A family of languages which is polynomial-time learnable from positive data in pitt's sense.
Erkki MäkinenPublished in: Int. J. Comput. Math. (1996)
Keyphrases
- positive data
- limit from positive data
- grammatical inference
- pattern languages
- languages from positive data
- tree languages
- regular languages
- positive and negative
- special case
- positive examples
- context free
- training data
- inductive inference
- noisy data
- finite automata
- accurate classification
- context free grammars
- decision rules
- learning algorithm