Abstract
In this paper, we developed a new algorithm, Gradient Weighted Matrix Factorisation (GWMF), for matrix factorisation. GWMF uses weights to focus the approximation in the matrix factorisation to the higher approxmation residual. Therefore, it improves the matrix factorisation accuracy and increases the speed of convergence. We also introduce a regularisation parameter to control overfitting. We applied our algorithm to a movie recommendation problem and GWMF performs better than ordinal gradient descent-based matrix factorisation (GMF) on Movielens dataset. GWMF converges faster than GMF and it guarantees lower root mean square error (RMSE) at earlier iterations on both training and testing.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
The size of the World Wide Web, http://www.worldwidewebsize.com/
The Size of Internet to Double Every 5 Years, http://www.labnol.org/internet/internet-size-to-double-every-5-years/6569/
Adomavicius, G., Tuzhilin, A.: Toward the Next Generation of Recommender Systems: A Survey of the State-of-the-Art and Possible Extensions. In: IEEE Transaction of Knowledge and Data Engineering, pp. 734–749 (2005)
Pazzani, M., Billsus, D.: Learning and Revising User Profiles: The identification of Interesting web sites. J. Mac. Lea. 27, 313–331 (1997)
Jannach, D., Friedrich, G.: Tutorial: Recommender Systems. In: Joint Conference on Artificial Intelligence, Barcelona (2011)
Salakhutdinov, R., Mnih, A.: Probabilistic Matrix Factorization. In: Neural Information Processing Systems (2008)
Hubert, L., Meulman, J., Heiser, W.: Two Purposes for Matrix Factorization: A Historical Appraisal. Society for Industrial and Applied Mathematics 42, 68–82 (2000)
Koren, Y., Bell, R.M., Volinsky, C.: Matrix Factorization Techniques for Recommender Systems. IEEE Computer 42(8), 30–37 (2009)
Salakhutdinov, R., Mnih, A.: Bayesian Probabilistic Matrix Factorization using Markov Chain Monte Carlo. In: 25th International Conference on Machine Learning (ICML), Helsinki (2008)
Zhang, S., Wang, W., Ford, J., Makedon, F.: Learning from Incomplete Ratings Using Non-negative Matrix Factorization. In: Proceedings of the SIAM International Conference on Data Mining, SDM (2006)
Gu, A., Zhou, J., Ding, C.: Collaborative Filtering: Weighted Nonnegative Matrix Factorization Incorporating User and Item Graphs. In: 10th SIAM International Conference on Data Mining (SDM), USA (2010)
Chen, G., Wang, F., Zhang, C.: Collaborative Filtering Using Orthogonal Nonnegative Matrix Trifactorization. J. of Inf. Pro. and Man. 45, 368–379 (2009)
Wu, M.: Collaborative Filtering via Ensembles of Matrix Factorizations. In: Proceedings of KDD Cup and Workshop, pp. 43–47 (2007)
Lee, D.D., Seung, H.S.: Algorithms for Non-Negative Matrix Factorization. In: Advances in Neural Information Processing, pp. 556–562 (2000)
Jin, R., Liu, Y., Si, L., Carbonell, J., Hauptmann, A.G.: A New Boosting Algorithm Using Input-Dependent Regularizer. In: Proceedings of the 20th International Conference on Machine Learning, Washington (2003)
MovieLens Data Sets, http://www.grouplens.org/node/73
Netflix Prize, http://www.netflixprize.com/
Gentle, J.: 6.8.1 Solutions that Minimize Other Norms of the Residuals. In: Matrix Algebra. Springer, New York (2007)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Chowdhury, N., Cai, X. (2013). GWMF: Gradient Weighted Matrix Factorisation for Recommender Systems. In: Ishikawa, Y., Li, J., Wang, W., Zhang, R., Zhang, W. (eds) Web Technologies and Applications. APWeb 2013. Lecture Notes in Computer Science, vol 7808. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-37401-2_73
Download citation
DOI: https://doi.org/10.1007/978-3-642-37401-2_73
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-37400-5
Online ISBN: 978-3-642-37401-2
eBook Packages: Computer ScienceComputer Science (R0)