Cautious Limit Learning

Vanja Doskoč, Timo Kötzing
; Proceedings of the 31st International Conference on Algorithmic Learning Theory, PMLR 117:251-276, 2020.

Abstract

We investigate language learning in the limit from text with various cautious learning restrictions. Learning is cautious if no hypothesis is a proper subset of a previous guess. While dealing with a seemingly natural learning behaviour, cautious learning does severely restrict explanatory (syntactic) learning power. To further understand why exactly this loss of learning power arises, Kötzing and Palenta (2016) introduced weakened versions of cautious learning and gave first partial results on their relation. In this paper, we aim to understand the restriction of cautious learning more fully. To this end we compare the known variants in a number of different settings, namely full-information and (partially) set-driven learning, paired either with the syntactic convergence restriction (explanatory learning) or the semantic convergence restriction (behaviourally correct learning). To do so, we make use of normal forms presented in Kötzing et al. (2017), most notably strongly locking and consistent learning. While strongly locking learners have been exploited when dealing with a variety of syntactic learning restrictions, we show how they can be beneficial in the semantic case as well. Furthermore, we expand the normal forms to a broader range of learning restrictions, including an answer to the open question of whether cautious learners can be assumed to be consistent, as stated in Kötzing et al. (2017).

Cite this Paper


BibTeX
@InProceedings{pmlr-v117-doskoc20a, title = {Cautious Limit Learning}, author = {Dosko\v{c}, Vanja and K\"{o}tzing, Timo}, booktitle = {Proceedings of the 31st International Conference on Algorithmic Learning Theory}, pages = {251--276}, year = {2020}, editor = {Aryeh Kontorovich and Gergely Neu}, volume = {117}, series = {Proceedings of Machine Learning Research}, address = {San Diego, California, USA}, month = {08 Feb--11 Feb}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v117/doskoc20a/doskoc20a.pdf}, url = {http://proceedings.mlr.press/v117/doskoc20a.html}, abstract = {We investigate language learning in the limit from text with various cautious learning restrictions. Learning is cautious if no hypothesis is a proper subset of a previous guess. While dealing with a seemingly natural learning behaviour, cautious learning does severely restrict explanatory (syntactic) learning power. To further understand why exactly this loss of learning power arises, Kötzing and Palenta (2016) introduced weakened versions of cautious learning and gave first partial results on their relation. In this paper, we aim to understand the restriction of cautious learning more fully. To this end we compare the known variants in a number of different settings, namely full-information and (partially) set-driven learning, paired either with the syntactic convergence restriction (explanatory learning) or the semantic convergence restriction (behaviourally correct learning). To do so, we make use of normal forms presented in Kötzing et al. (2017), most notably strongly locking and consistent learning. While strongly locking learners have been exploited when dealing with a variety of syntactic learning restrictions, we show how they can be beneficial in the semantic case as well. Furthermore, we expand the normal forms to a broader range of learning restrictions, including an answer to the open question of whether cautious learners can be assumed to be consistent, as stated in Kötzing et al. (2017).} }
Endnote
%0 Conference Paper %T Cautious Limit Learning %A Vanja Doskoč %A Timo Kötzing %B Proceedings of the 31st International Conference on Algorithmic Learning Theory %C Proceedings of Machine Learning Research %D 2020 %E Aryeh Kontorovich %E Gergely Neu %F pmlr-v117-doskoc20a %I PMLR %J Proceedings of Machine Learning Research %P 251--276 %U http://proceedings.mlr.press %V 117 %W PMLR %X We investigate language learning in the limit from text with various cautious learning restrictions. Learning is cautious if no hypothesis is a proper subset of a previous guess. While dealing with a seemingly natural learning behaviour, cautious learning does severely restrict explanatory (syntactic) learning power. To further understand why exactly this loss of learning power arises, Kötzing and Palenta (2016) introduced weakened versions of cautious learning and gave first partial results on their relation. In this paper, we aim to understand the restriction of cautious learning more fully. To this end we compare the known variants in a number of different settings, namely full-information and (partially) set-driven learning, paired either with the syntactic convergence restriction (explanatory learning) or the semantic convergence restriction (behaviourally correct learning). To do so, we make use of normal forms presented in Kötzing et al. (2017), most notably strongly locking and consistent learning. While strongly locking learners have been exploited when dealing with a variety of syntactic learning restrictions, we show how they can be beneficial in the semantic case as well. Furthermore, we expand the normal forms to a broader range of learning restrictions, including an answer to the open question of whether cautious learners can be assumed to be consistent, as stated in Kötzing et al. (2017).
APA
Doskoč, V. & Kötzing, T.. (2020). Cautious Limit Learning. Proceedings of the 31st International Conference on Algorithmic Learning Theory, in PMLR 117:251-276

Related Material