TY - GEN
T1 - Learning commutative regular languages
AU - Gómez, Antonio Cano
AU - Álvarez, Gloria I.
N1 - Funding Information:
Work partially supported by Ministerio de Educación y Ciencia under project TIN2007-60760.
PY - 2008
Y1 - 2008
N2 - In this article we study the inference of commutative regular languages. We first show that commutative regular languages are not inferable from positive samples, and then we study the possible improvement of inference from positive and negative samples. We propose a polynomial algorithm to infer commutative regular languages from positive and negative samples, and we show, from experimental results, that far from being a theoretical algorithm, it produces very high recognition rates in comparison with classical inference algorithms.
AB - In this article we study the inference of commutative regular languages. We first show that commutative regular languages are not inferable from positive samples, and then we study the possible improvement of inference from positive and negative samples. We propose a polynomial algorithm to infer commutative regular languages from positive and negative samples, and we show, from experimental results, that far from being a theoretical algorithm, it produces very high recognition rates in comparison with classical inference algorithms.
UR - http://www.scopus.com/inward/record.url?scp=56649118829&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-88009-7_6
DO - 10.1007/978-3-540-88009-7_6
M3 - Conference contribution
AN - SCOPUS:56649118829
SN - 3540880089
SN - 9783540880080
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 71
EP - 83
BT - Grammatical Inference
T2 - 9th International Colloquium on Grammatical Inference, ICGI 2008
Y2 - 22 September 2008 through 24 September 2008
ER -