Polynomial Matrix Completion for Missing Data Imputation and Transductive Learning

被引:0
作者
Fan, Jicong [1 ]
Zhang, Yuqian [1 ]
Udell, Madeleine [1 ]
机构
[1] Cornell Univ, Ithaca, NY 14853 USA
来源
THIRTY-FOURTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THE THIRTY-SECOND INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE AND THE TENTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE | 2020年 / 34卷
关键词
NORM; DIMENSIONALITY; MINIMIZATION;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
This paper develops new methods to recover the missing entries of a high-rank or even full-rank matrix when the intrinsic dimension of the data is low compared to the ambient dimension. Specifically, we assume that the columns of a matrix are generated by polynomials acting on a low-dimensional intrinsic variable, and wish to recover the missing entries under this assumption. We show that we can identify the complete matrix of minimum intrinsic dimension by minimizing the rank of the matrix in a high dimensional feature space. We develop a new formulation of the resulting problem using the kernel trick together with a new relaxation of the rank objective, and propose an efficient optimization method. We also show how to use our methods to complete data drawn from multiple nonlinear manifolds. Comparative studies on synthetic data, subspace clustering with missing data, motion capture data recovery, and transductive learning verify the superiority of our methods over the state-of-the-art.
引用
收藏
页码:3842 / 3849
页数:8
相关论文
共 36 条
[1]  
[Anonymous], 2016, PROC CVPR IEEE, DOI DOI 10.1109/CVPR.2016.566
[2]   Exact Matrix Completion via Convex Optimization [J].
Candes, Emmanuel J. ;
Recht, Benjamin .
FOUNDATIONS OF COMPUTATIONAL MATHEMATICS, 2009, 9 (06) :717-772
[3]  
Elhamifar E, 2016, ADV NEUR IN, V29
[4]   Sparse Subspace Clustering: Algorithm, Theory, and Applications [J].
Elhamifar, Ehsan ;
Vidal, Rene .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2013, 35 (11) :2765-2781
[5]  
Eriksson Brian., 2011, CoRR
[6]  
Fan J., 2018, IEEE T BIG DATA
[7]   Non-linear matrix completion [J].
Fan, Jicong ;
Chow, Tommy W. S. .
PATTERN RECOGNITION, 2018, 77 :378-394
[8]   Matrix completion by deep matrix factorization [J].
Fan, Jicong ;
Cheng, Jieyu .
NEURAL NETWORKS, 2018, 98 :34-41
[9]   Matrix completion by least-square, low-rank, and sparse self-representations [J].
Fan, Jicong ;
Chow, Tommy W. S. .
PATTERN RECOGNITION, 2017, 71 :290-305
[10]   Log-det heuristic for matrix rank minimization with applications to Hankel and Euclidean distance matrices [J].
Fazel, M ;
Hindi, H ;
Boyd, SP .
PROCEEDINGS OF THE 2003 AMERICAN CONTROL CONFERENCE, VOLS 1-6, 2003, :2156-2162