Programming by Feedback

Marc Schoenauer, Riad Akrour, Michele Sebag, Jean-Christophe Souplet
Proceedings of the 31st International Conference on Machine Learning, PMLR 32(2):1503-1511, 2014.

Abstract

This paper advocates a new ML-based programming framework, called Programming by Feedback (PF), which involves a sequence of interactions between the active computer and the user. The latter only provides preference judgments on pairs of solutions supplied by the active computer. The active computer involves two components: the learning component estimates the user’s utility function and accounts for the user’s (possibly limited) competence; the optimization component explores the search space and returns the most appropriate candidate solution. A proof of principle of the approach is proposed, showing that PF requires a handful of interactions in order to solve some discrete and continuous benchmark problems.

Cite this Paper

BibTeX
@InProceedings{pmlr-v32-schoenauer14, title = {Programming by Feedback}, author = {Schoenauer, Marc and Akrour, Riad and Sebag, Michele and Souplet, Jean-Christophe}, booktitle = {Proceedings of the 31st International Conference on Machine Learning}, pages = {1503--1511}, year = {2014}, editor = {Xing, Eric P. and Jebara, Tony}, volume = {32}, number = {2}, series = {Proceedings of Machine Learning Research}, address = {Bejing, China}, month = {22--24 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v32/schoenauer14.pdf}, url = {https://proceedings.mlr.press/v32/schoenauer14.html}, abstract = {This paper advocates a new ML-based programming framework, called Programming by Feedback (PF), which involves a sequence of interactions between the active computer and the user. The latter only provides preference judgments on pairs of solutions supplied by the active computer. The active computer involves two components: the learning component estimates the user’s utility function and accounts for the user’s (possibly limited) competence; the optimization component explores the search space and returns the most appropriate candidate solution. A proof of principle of the approach is proposed, showing that PF requires a handful of interactions in order to solve some discrete and continuous benchmark problems.} }
Endnote
%0 Conference Paper %T Programming by Feedback %A Marc Schoenauer %A Riad Akrour %A Michele Sebag %A Jean-Christophe Souplet %B Proceedings of the 31st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2014 %E Eric P. Xing %E Tony Jebara %F pmlr-v32-schoenauer14 %I PMLR %P 1503--1511 %U https://proceedings.mlr.press/v32/schoenauer14.html %V 32 %N 2 %X This paper advocates a new ML-based programming framework, called Programming by Feedback (PF), which involves a sequence of interactions between the active computer and the user. The latter only provides preference judgments on pairs of solutions supplied by the active computer. The active computer involves two components: the learning component estimates the user’s utility function and accounts for the user’s (possibly limited) competence; the optimization component explores the search space and returns the most appropriate candidate solution. A proof of principle of the approach is proposed, showing that PF requires a handful of interactions in order to solve some discrete and continuous benchmark problems.
RIS
TY - CPAPER TI - Programming by Feedback AU - Marc Schoenauer AU - Riad Akrour AU - Michele Sebag AU - Jean-Christophe Souplet BT - Proceedings of the 31st International Conference on Machine Learning DA - 2014/06/18 ED - Eric P. Xing ED - Tony Jebara ID - pmlr-v32-schoenauer14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 32 IS - 2 SP - 1503 EP - 1511 L1 - http://proceedings.mlr.press/v32/schoenauer14.pdf UR - https://proceedings.mlr.press/v32/schoenauer14.html AB - This paper advocates a new ML-based programming framework, called Programming by Feedback (PF), which involves a sequence of interactions between the active computer and the user. The latter only provides preference judgments on pairs of solutions supplied by the active computer. The active computer involves two components: the learning component estimates the user’s utility function and accounts for the user’s (possibly limited) competence; the optimization component explores the search space and returns the most appropriate candidate solution. A proof of principle of the approach is proposed, showing that PF requires a handful of interactions in order to solve some discrete and continuous benchmark problems. ER -
APA
Schoenauer, M., Akrour, R., Sebag, M. & Souplet, J.. (2014). Programming by Feedback. Proceedings of the 31st International Conference on Machine Learning, in Proceedings of Machine Learning Research 32(2):1503-1511 Available from https://proceedings.mlr.press/v32/schoenauer14.html.

Related Material