Improved SAT Models for NFA Learning

Lardeux, Frédéric; Monfroy, Éric

doi:10.1007/978-3-030-85672-4_20

Cited by 5 publications

(3 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [12], we proposed two strategies based on metaheuristics for optimizing models, and consequently for optimizing the split of each word of S. For both of them, the search space corresponds to all the hybrid models, i.e., all the possible split for all words of S. The fitness we used is:…”

Section: Previous Modelsmentioning

confidence: 99%

“…This task is called modeling, the result (variables and constraints) is a model, and a model together with some data forms an instance. In our case, the model is a NFA inference model, the data are S and k. For example, an INLP (Integer Non Linear Programming) model for inferring NFA is given in [10], in [2] a SAT (the propositional satisfiability problem [11]) model is given, and in [12], [1] several improvements of the SAT model are given.…”

Section: Introductionmentioning

confidence: 99%

“…Whereas some approaches focus on improving the solver, such as specific strategies [13], or parallel techniques [2] for optimizing k, we have an othogonal approach that consists in improving the model and using a standard solver. In [12], we studied complex data structures to generate smaller instances, and in [1], we proposed various models, and we use metaheuristics to optimize instance sizes. In this paper, we propose four new models for the inference of NFA based on the optimization of the splitting of words into a prefix and a suffix.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Optimized models and symmetry breaking for the NFA inference problem

Lardeux

Monfroy

2021

2021 IEEE 33rd International Conference on Tools With Artificial Intelligence (ICTAI)

Self Cite

View full text Add to dashboard Cite

Grammatical inference is concerned with the study of algorithms for learning automata and grammars from words. We propose some models for learning Nondeterministic Finite Automaton (NFA) of size k from samples of words of the language and words not belonging to the language we want to describe. To this end, we formulate the problem as a SAT model trying to reduce the size of generated SAT instances.We propose new models to generate even smaller SAT instances. We also suggest some techniques for breaking some symmetries, hence reducing the search space, and consequently, speeding-up solving. We also achieved some experimental comparisons and we analyze our various model improvements and over-constraint propositions. Compared to [1], our models are easier and faster to solve. Compare to the parallel solver of [2], we find some new bounds for some instances for which the minimal size of NFA is not known yet.

show abstract

Section: Previous Modelsmentioning

confidence: 99%