Learning Recursive Automata from Positive Examples
نویسنده
چکیده
In this theoretical paper, we compare the “classical” learning techniques used to infer regular grammars from positive examples with the ones used to infer categorial grammars. To this aim, we first study how to translate finite state automata into categorial grammars and back. We then show that the generalization operators employed in both domains can be compared, and that their result can always be represented by generalized automata, called “recursive automata”. The relation between these generalized automata and categorial grammars is studied in detail. Finally, new learnable subclasses of categorial grammars are defined, for which learning from strings is nearly not more expensive than from structures. RÉSUMÉ. Dans cet article théorique, nous proposons de comparer les techniques “classiques” employées en inférence grammaticale de langages réguliers par exemples positifs avec celles employées pour l’inférence de grammaires catégorielles. Pour cela, nous commençons par étudier les traductions réciproques entre automates finis et grammaires catégorielles. Nous montrons ensuite que les opérateurs de généralisation utilisés dans chacun des domaines sont comparables, et que le résultat de leur application peut toujours se représenter à l’aide d’automates généralisés appelés “récursifs”. Les liens entre ces automates généralisés et les grammaires catégorielles sont étudiés en détail. Enfin, nous exhibons de nouvelles sous-classes apprenables de grammaires catégorielles pour lesquelles l’apprentissage à partir de textes n’est presque pas plus coûteux que l’apprentissage à partir de structures.
منابع مشابه
Theory-Guideed Induction of Logic Programs by Inference of Regular Languages
Previous resolution-based approaches to theory-guided induction of logic programs produce hypotheses in the form of a set of resolvents of a theory, where the resolvents represent allowed sequences of resolution steps for the initial theory. There are, however, many characterizations of allowed sequences of resolution steps that cannot be expressed by a set of resolvents. One approach to this p...
متن کاملرمزنگاری تصاویر با استفاده از اتوماتای سلولی برگشت پذیر
In this paper, a new structure for image encryption using recursive cellular automatais presented. The image encryption contains three recursive cellular automata in three steps, individually. At the first step, the image is blocked and the pixels are substituted. In the next step, pixels are scrambledby the second cellular automata and at the last step, the blocks are attachedtogether and the ...
متن کاملLearning Simple Recursive Concepts by Discovering Missing Examples
Here we introduce a system, SmartPlus, which learns recursive concepts from a small incomplete training set the members of which all lie on non-intersecting resolution path with respect to the target recursive theory and involve dierent constants. Unlike recent approaches to this learning problem, our method is based upon discovering the missing examples (facts corresponding to the rst recursiv...
متن کاملLearning Restarting Automata by Genetic Algorithms
Restarting automaton is a special type of a linear bounded automaton designed for modelling the so-called analysis by reduction. We use genetic algorithms to learn restarting automata to recognize languages according to input consisting of sets of positive and negative examples of words from the language together with positive and negative examples of simplifications.
متن کاملLearning Analysis by Reduction from Positive Data
Analysis by reduction is a linguistically motivated method for checking correctness of a sentence. It can be modelled by restarting automata. In this paper we propose a method for learning of restarting automata which are strictly locally testable (SLT-R-automata). The method is based on the concept of identification in the limit from positive examples only. Also we characterize the class of la...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید
ثبت ناماگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید
ورودعنوان ژورنال:
- Revue d'Intelligence Artificielle
دوره 20 شماره
صفحات -
تاریخ انتشار 2006