Box Constrained Low-rank Matrix Approximation with Missing Values

Manami Tatsukawa, Mirai Tanaka

Abstract

In this paper, we propose a new low-rank matrix approximation model for completing a matrix with missing values. Our proposed model contains a box constraint that arises from the context of collaborative filtering. Although it is unfortunately NP-hard to solve our model with high accuracy, we can construct a practical algorithm to obtain a stationary point. Our proposed algorithm is based on alternating minimization and converges to a stationary point under a mild assumption.

References

Download


Paper Citation


in Harvard Style

Tatsukawa M. and Tanaka M. (2018). Box Constrained Low-rank Matrix Approximation with Missing Values.In Proceedings of the 7th International Conference on Operations Research and Enterprise Systems - Volume 1: ICORES, ISBN 978-989-758-285-1, pages 78-84. DOI: 10.5220/0006612100780084


in Bibtex Style

@conference{icores18,
author={Manami Tatsukawa and Mirai Tanaka},
title={Box Constrained Low-rank Matrix Approximation with Missing Values},
booktitle={Proceedings of the 7th International Conference on Operations Research and Enterprise Systems - Volume 1: ICORES,},
year={2018},
pages={78-84},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006612100780084},
isbn={978-989-758-285-1},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 7th International Conference on Operations Research and Enterprise Systems - Volume 1: ICORES,
TI - Box Constrained Low-rank Matrix Approximation with Missing Values
SN - 978-989-758-285-1
AU - Tatsukawa M.
AU - Tanaka M.
PY - 2018
SP - 78
EP - 84
DO - 10.5220/0006612100780084