Bounded Off-Policy Evaluation with Missing Data for Course Recommendation and Curriculum Design

William Hoiles, Mihaela Schaar
Proceedings of The 33rd International Conference on Machine Learning, PMLR 48:1596-1604, 2016.

Abstract

Successfully recommending personalized course schedules is a difficult problem given the diversity of students knowledge, learning behaviour, and goals. This paper presents personalized course recommendation and curriculum design algorithms that exploit logged student data. The algorithms are based on the regression estimator for contextual multi-armed bandits with a penalized variance term. Guarantees on the predictive performance of the algorithms are provided using empirical Bernstein bounds. We also provide guidelines for including expert domain knowledge into the recommendations. Using undergraduate engineering logged data from a post-secondary institution we illustrate the performance of these algorithms.

Cite this Paper


BibTeX
@InProceedings{pmlr-v48-hoiles16, title = {Bounded Off-Policy Evaluation with Missing Data for Course Recommendation and Curriculum Design}, author = {Hoiles, William and Schaar, Mihaela}, booktitle = {Proceedings of The 33rd International Conference on Machine Learning}, pages = {1596--1604}, year = {2016}, editor = {Balcan, Maria Florina and Weinberger, Kilian Q.}, volume = {48}, series = {Proceedings of Machine Learning Research}, address = {New York, New York, USA}, month = {20--22 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v48/hoiles16.pdf}, url = {https://proceedings.mlr.press/v48/hoiles16.html}, abstract = {Successfully recommending personalized course schedules is a difficult problem given the diversity of students knowledge, learning behaviour, and goals. This paper presents personalized course recommendation and curriculum design algorithms that exploit logged student data. The algorithms are based on the regression estimator for contextual multi-armed bandits with a penalized variance term. Guarantees on the predictive performance of the algorithms are provided using empirical Bernstein bounds. We also provide guidelines for including expert domain knowledge into the recommendations. Using undergraduate engineering logged data from a post-secondary institution we illustrate the performance of these algorithms.} }
Endnote
%0 Conference Paper %T Bounded Off-Policy Evaluation with Missing Data for Course Recommendation and Curriculum Design %A William Hoiles %A Mihaela Schaar %B Proceedings of The 33rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2016 %E Maria Florina Balcan %E Kilian Q. Weinberger %F pmlr-v48-hoiles16 %I PMLR %P 1596--1604 %U https://proceedings.mlr.press/v48/hoiles16.html %V 48 %X Successfully recommending personalized course schedules is a difficult problem given the diversity of students knowledge, learning behaviour, and goals. This paper presents personalized course recommendation and curriculum design algorithms that exploit logged student data. The algorithms are based on the regression estimator for contextual multi-armed bandits with a penalized variance term. Guarantees on the predictive performance of the algorithms are provided using empirical Bernstein bounds. We also provide guidelines for including expert domain knowledge into the recommendations. Using undergraduate engineering logged data from a post-secondary institution we illustrate the performance of these algorithms.
RIS
TY - CPAPER TI - Bounded Off-Policy Evaluation with Missing Data for Course Recommendation and Curriculum Design AU - William Hoiles AU - Mihaela Schaar BT - Proceedings of The 33rd International Conference on Machine Learning DA - 2016/06/11 ED - Maria Florina Balcan ED - Kilian Q. Weinberger ID - pmlr-v48-hoiles16 PB - PMLR DP - Proceedings of Machine Learning Research VL - 48 SP - 1596 EP - 1604 L1 - http://proceedings.mlr.press/v48/hoiles16.pdf UR - https://proceedings.mlr.press/v48/hoiles16.html AB - Successfully recommending personalized course schedules is a difficult problem given the diversity of students knowledge, learning behaviour, and goals. This paper presents personalized course recommendation and curriculum design algorithms that exploit logged student data. The algorithms are based on the regression estimator for contextual multi-armed bandits with a penalized variance term. Guarantees on the predictive performance of the algorithms are provided using empirical Bernstein bounds. We also provide guidelines for including expert domain knowledge into the recommendations. Using undergraduate engineering logged data from a post-secondary institution we illustrate the performance of these algorithms. ER -
APA
Hoiles, W. & Schaar, M.. (2016). Bounded Off-Policy Evaluation with Missing Data for Course Recommendation and Curriculum Design. Proceedings of The 33rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 48:1596-1604 Available from https://proceedings.mlr.press/v48/hoiles16.html.

Related Material