Inductive Inference from Positive Data Is Powerful
Abstract
ABSTRACT Inductive inference from positive data is shown to be remarkably powerful using a framework of elementary formal system. An elementary formal system, EFS for short, is a kind of logic program on σ+ consisting of finitely many axioms. Any context-sensitive language is definable by a restricted EFS, called length-bounded EFS. Length-bounded EFS's with at most n axioms are considered to show that inductive inference from positive data works successfully for their models as well as for their languages. From this it follows that any class of usual logic programs, like Prolog programs, corresponding to length-bounded EFS's can be inferred from positive facts.
Cite
Text
Shinohara. "Inductive Inference from Positive Data Is Powerful." Annual Conference on Computational Learning Theory, 1990. doi:10.1016/B978-1-55860-146-8.50010-2Markdown
[Shinohara. "Inductive Inference from Positive Data Is Powerful." Annual Conference on Computational Learning Theory, 1990.](https://mlanthology.org/colt/1990/shinohara1990colt-inductive/) doi:10.1016/B978-1-55860-146-8.50010-2BibTeX
@inproceedings{shinohara1990colt-inductive,
title = {{Inductive Inference from Positive Data Is Powerful}},
author = {Shinohara, Takeshi},
booktitle = {Annual Conference on Computational Learning Theory},
year = {1990},
pages = {97-110},
doi = {10.1016/B978-1-55860-146-8.50010-2},
url = {https://mlanthology.org/colt/1990/shinohara1990colt-inductive/}
}