A non-IID Framework for Collaborative Filtering with Restricted Boltzmann Machines

Kostadin Georgiev, Preslav Nakov
; Proceedings of the 30th International Conference on Machine Learning, PMLR 28(3):1148-1156, 2013.

Abstract

We propose a framework for collaborative filtering based on Restricted Boltzmann Machines (RBM), which extends previous RBM-based approaches in several important directions. First, while previous RBM research has focused on modeling the correlation between item ratings, we model both user-user and item-item correlations in a unified hybrid non-IID framework. We further use real values in the visible layer as opposed to multinomial variables, thus taking advantage of the natural order between user-item ratings. Finally, we explore the potential of combining the original training data with data generated by the RBM-based model itself in a bootstrapping fashion. The evaluation on two MovieLens datasets (with 100K and 1M user-item ratings, respectively), shows that our RBM model rivals the best previously-proposed approaches.

Cite this Paper


BibTeX
@InProceedings{pmlr-v28-georgiev13, title = {A non-IID Framework for Collaborative Filtering with Restricted Boltzmann Machines}, author = {Kostadin Georgiev and Preslav Nakov}, booktitle = {Proceedings of the 30th International Conference on Machine Learning}, pages = {1148--1156}, year = {2013}, editor = {Sanjoy Dasgupta and David McAllester}, volume = {28}, number = {3}, series = {Proceedings of Machine Learning Research}, address = {Atlanta, Georgia, USA}, month = {17--19 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v28/georgiev13.pdf}, url = {http://proceedings.mlr.press/v28/georgiev13.html}, abstract = {We propose a framework for collaborative filtering based on Restricted Boltzmann Machines (RBM), which extends previous RBM-based approaches in several important directions. First, while previous RBM research has focused on modeling the correlation between item ratings, we model both user-user and item-item correlations in a unified hybrid non-IID framework. We further use real values in the visible layer as opposed to multinomial variables, thus taking advantage of the natural order between user-item ratings. Finally, we explore the potential of combining the original training data with data generated by the RBM-based model itself in a bootstrapping fashion. The evaluation on two MovieLens datasets (with 100K and 1M user-item ratings, respectively), shows that our RBM model rivals the best previously-proposed approaches. } }
Endnote
%0 Conference Paper %T A non-IID Framework for Collaborative Filtering with Restricted Boltzmann Machines %A Kostadin Georgiev %A Preslav Nakov %B Proceedings of the 30th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2013 %E Sanjoy Dasgupta %E David McAllester %F pmlr-v28-georgiev13 %I PMLR %J Proceedings of Machine Learning Research %P 1148--1156 %U http://proceedings.mlr.press %V 28 %N 3 %W PMLR %X We propose a framework for collaborative filtering based on Restricted Boltzmann Machines (RBM), which extends previous RBM-based approaches in several important directions. First, while previous RBM research has focused on modeling the correlation between item ratings, we model both user-user and item-item correlations in a unified hybrid non-IID framework. We further use real values in the visible layer as opposed to multinomial variables, thus taking advantage of the natural order between user-item ratings. Finally, we explore the potential of combining the original training data with data generated by the RBM-based model itself in a bootstrapping fashion. The evaluation on two MovieLens datasets (with 100K and 1M user-item ratings, respectively), shows that our RBM model rivals the best previously-proposed approaches.
RIS
TY - CPAPER TI - A non-IID Framework for Collaborative Filtering with Restricted Boltzmann Machines AU - Kostadin Georgiev AU - Preslav Nakov BT - Proceedings of the 30th International Conference on Machine Learning PY - 2013/02/13 DA - 2013/02/13 ED - Sanjoy Dasgupta ED - David McAllester ID - pmlr-v28-georgiev13 PB - PMLR SP - 1148 DP - PMLR EP - 1156 L1 - http://proceedings.mlr.press/v28/georgiev13.pdf UR - http://proceedings.mlr.press/v28/georgiev13.html AB - We propose a framework for collaborative filtering based on Restricted Boltzmann Machines (RBM), which extends previous RBM-based approaches in several important directions. First, while previous RBM research has focused on modeling the correlation between item ratings, we model both user-user and item-item correlations in a unified hybrid non-IID framework. We further use real values in the visible layer as opposed to multinomial variables, thus taking advantage of the natural order between user-item ratings. Finally, we explore the potential of combining the original training data with data generated by the RBM-based model itself in a bootstrapping fashion. The evaluation on two MovieLens datasets (with 100K and 1M user-item ratings, respectively), shows that our RBM model rivals the best previously-proposed approaches. ER -
APA
Georgiev, K. & Nakov, P.. (2013). A non-IID Framework for Collaborative Filtering with Restricted Boltzmann Machines. Proceedings of the 30th International Conference on Machine Learning, in PMLR 28(3):1148-1156

Related Material