Pattern languages are not learnable
COLT '90 Proceedings of the third annual workshop on Computational learning theory
A machine discovery from amino acid sequences by decision trees over regular patterns
Selected papers of international conference on Fifth generation computer systems 92
An introduction to computational learning theory
An introduction to computational learning theory
Characterizations of monotonic and dual monotonic language learning
Information and Computation
Learning one-variable pattern languages in linear average time
COLT' 98 Proceedings of the eleventh annual conference on Computational learning theory
Introduction To Automata Theory, Languages, And Computation
Introduction To Automata Theory, Languages, And Computation
Inclusion is Undecidable for Pattern Languages
ICALP '93 Proceedings of the 20th International Colloquium on Automata, Languages and Programming
Algorithmic Learning for Knowledge-Based Systems, GOSLER Final Report
A Guided Tour Across the Boundaries of Learning Recursive Languages
Algorithmic Learning for Knowledge-Based Systems, GOSLER Final Report
ALT '97 Proceedings of the 8th International Conference on Algorithmic Learning Theory
Monotone Extensions of Boolean Data Sets
ALT '97 Proceedings of the 8th International Conference on Algorithmic Learning Theory
Hi-index | 0.00 |
The present paper deals with the problem of finding a consistent one-variable pattern from incomplete positive and negative examples. The studied problems are called an extension e, a consistent extension ce and a robust extension re, respectively. Problem e corresponds to the ordinary problem to decide whether there exists a one-variable pattern that is consistent with the given positive and negative examples. As for the other problems, an example string is allowed to contain some unsettled symbols that can potentially match with every constant symbol. For the problem CE, one has to decide whether there exists a suitable assignment for these unsettled symbols as well as a one-variable pattern consistent with the examples with respect to the assignment chosen. Problem RE is the universal version of problem ce, i.e., now one has to decide whether there exists a one-variable pattern that is consistent with the examples under every assignment for the unsettled symbols. The decision problems defined are closely connected to the learnability of one-variable pattern languages from positive and negative examples. The computational complexity of the decision problems defined above is studied. In particular, it shown that RE is NP-complete.