Statistical mechanics of low-rank tensor decomposition

被引:7
作者
Kadmon, Jonathan [1 ]
Ganguli, Surya [2 ,3 ]
机构
[1] Stanford Univ, Dept Appl Phys, Stanford, CA 94305 USA
[2] Stanford Univ, Dept Appl Phys, Mountain View, CA USA
[3] Google Brain, Mountain View, CA USA
基金
美国国家卫生研究院;
关键词
machine learning; MODEL;
D O I
10.1088/1742-5468/ab3216
中图分类号
O3 [力学];
学科分类号
08 ; 0801 ;
摘要
Often, large, high-dimensional datasets collected across multiple modalities can be organized as a higher-order tensor. Low-rank tensor decomposition then arises as a powerful and widely used tool to discover simple low-dimensional structures underlying such data. However, we currently lack a theoretical understanding of the algorithmic behavior of low-rank tensor decompositions. We derive Bayesian approximate message passing (AMP) algorithms for recovering arbitrarily shaped low-rank tensors buried within noise, and we employ dynamic mean field theory to precisely characterize their performance. Our theory reveals the existence of phase transitions between easy, hard and impossible inference regimes, and displays an excellent match with simulations. Moreover it reveals several qualitative surprises compared to the behavior of symmetric, cubic tensor decomposition. Finally, we compare our AMP algorithm to the most commonly used algorithm, alternating least squares (ALS), and demonstrate that AMP significantly outperforms ALS in the presence of noise.
引用
收藏
页数:15
相关论文
共 38 条
[1]   Multiway analysis of epilepsy tensors [J].
Acar, Evrim ;
Aykut-Bingol, Canan ;
Bingol, Haluk ;
Bro, Rasmus ;
Yener, Buelent .
BIOINFORMATICS, 2007, 23 (13) :I10-I18
[2]  
Advani M., 2016, Advances in Neural Information Processing Systems, P3378
[3]   Statistical Mechanics of Optimal Convex Inference in High Dimensions [J].
Advani, Madhu ;
Ganguli, Surya .
PHYSICAL REVIEW X, 2016, 6 (03)
[4]   Statistical mechanics of complex neural systems and high dimensional data [J].
Advani, Madhu ;
Lahiri, Subhaneil ;
Ganguli, Surya .
JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT, 2013,
[5]  
Anandkumar A, 2014, J MACH LEARN RES, V15, P2773
[6]  
[Anonymous], 2001, STAT PHYS SPIN GLASS
[7]  
[Anonymous], 2009, Advances in Neural Information Processing Systems
[8]  
Barbier J, 2017, ANN ALLERTON CONF, P1056, DOI 10.1109/ALLERTON.2017.8262854
[9]   The Dynamics of Message Passing on Dense Graphs, with Applications to Compressed Sensing [J].
Bayati, Mohsen ;
Montanari, Andrea .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2011, 57 (02) :764-785
[10]  
Bethe Hans A., 1935, Proceedings of the Royal Society A, V150, P552, DOI DOI 10.1098/RSPA.1935.0122