Cold-start Active Learning with Robust Ordinal Matrix Factorization

Neil Houlsby, Jose Miguel Hernandez-Lobato, Zoubin Ghahramani
Proceedings of the 31st International Conference on Machine Learning, PMLR 32(2):766-774, 2014.

Abstract

We present a new matrix factorization model for rating data and a corresponding active learning strategy to address the cold-start problem. Cold-start is one of the most challenging tasks for recommender systems: what to recommend with new users or items for which one has little or no data. An approach is to use active learning to collect the most useful initial ratings. However, the performance of active learning depends strongly upon having accurate estimates of i) the uncertainty in model parameters and ii) the intrinsic noisiness of the data. To achieve these estimates we propose a heteroskedastic Bayesian model for ordinal matrix factorization. We also present a computationally efficient framework for Bayesian active learning with this type of complex probabilistic model. This algorithm successfully distinguishes between informative and noisy data points. Our model yields state-of-the-art predictive performance and, coupled with our active learning strategy, enables us to gain useful information in the cold-start setting from the very first active sample.

Cite this Paper

BibTeX
@InProceedings{pmlr-v32-houlsby14, title = {Cold-start Active Learning with Robust Ordinal Matrix Factorization}, author = {Houlsby, Neil and Hernandez-Lobato, Jose Miguel and Ghahramani, Zoubin}, booktitle = {Proceedings of the 31st International Conference on Machine Learning}, pages = {766--774}, 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/houlsby14.pdf}, url = {https://proceedings.mlr.press/v32/houlsby14.html}, abstract = {We present a new matrix factorization model for rating data and a corresponding active learning strategy to address the cold-start problem. Cold-start is one of the most challenging tasks for recommender systems: what to recommend with new users or items for which one has little or no data. An approach is to use active learning to collect the most useful initial ratings. However, the performance of active learning depends strongly upon having accurate estimates of i) the uncertainty in model parameters and ii) the intrinsic noisiness of the data. To achieve these estimates we propose a heteroskedastic Bayesian model for ordinal matrix factorization. We also present a computationally efficient framework for Bayesian active learning with this type of complex probabilistic model. This algorithm successfully distinguishes between informative and noisy data points. Our model yields state-of-the-art predictive performance and, coupled with our active learning strategy, enables us to gain useful information in the cold-start setting from the very first active sample.} }
Endnote
%0 Conference Paper %T Cold-start Active Learning with Robust Ordinal Matrix Factorization %A Neil Houlsby %A Jose Miguel Hernandez-Lobato %A Zoubin Ghahramani %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-houlsby14 %I PMLR %P 766--774 %U https://proceedings.mlr.press/v32/houlsby14.html %V 32 %N 2 %X We present a new matrix factorization model for rating data and a corresponding active learning strategy to address the cold-start problem. Cold-start is one of the most challenging tasks for recommender systems: what to recommend with new users or items for which one has little or no data. An approach is to use active learning to collect the most useful initial ratings. However, the performance of active learning depends strongly upon having accurate estimates of i) the uncertainty in model parameters and ii) the intrinsic noisiness of the data. To achieve these estimates we propose a heteroskedastic Bayesian model for ordinal matrix factorization. We also present a computationally efficient framework for Bayesian active learning with this type of complex probabilistic model. This algorithm successfully distinguishes between informative and noisy data points. Our model yields state-of-the-art predictive performance and, coupled with our active learning strategy, enables us to gain useful information in the cold-start setting from the very first active sample.
RIS
TY - CPAPER TI - Cold-start Active Learning with Robust Ordinal Matrix Factorization AU - Neil Houlsby AU - Jose Miguel Hernandez-Lobato AU - Zoubin Ghahramani BT - Proceedings of the 31st International Conference on Machine Learning DA - 2014/06/18 ED - Eric P. Xing ED - Tony Jebara ID - pmlr-v32-houlsby14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 32 IS - 2 SP - 766 EP - 774 L1 - http://proceedings.mlr.press/v32/houlsby14.pdf UR - https://proceedings.mlr.press/v32/houlsby14.html AB - We present a new matrix factorization model for rating data and a corresponding active learning strategy to address the cold-start problem. Cold-start is one of the most challenging tasks for recommender systems: what to recommend with new users or items for which one has little or no data. An approach is to use active learning to collect the most useful initial ratings. However, the performance of active learning depends strongly upon having accurate estimates of i) the uncertainty in model parameters and ii) the intrinsic noisiness of the data. To achieve these estimates we propose a heteroskedastic Bayesian model for ordinal matrix factorization. We also present a computationally efficient framework for Bayesian active learning with this type of complex probabilistic model. This algorithm successfully distinguishes between informative and noisy data points. Our model yields state-of-the-art predictive performance and, coupled with our active learning strategy, enables us to gain useful information in the cold-start setting from the very first active sample. ER -
APA
Houlsby, N., Hernandez-Lobato, J.M. & Ghahramani, Z.. (2014). Cold-start Active Learning with Robust Ordinal Matrix Factorization. Proceedings of the 31st International Conference on Machine Learning, in Proceedings of Machine Learning Research 32(2):766-774 Available from https://proceedings.mlr.press/v32/houlsby14.html.

Related Material