Learning from Binary Multiway Data: Probabilistic Tensor Decomposition and its Statistical Optimality

We consider the problem of decomposing a higher-order tensor with binary entries. Such data problems arise frequently in applications such as neuroimaging, recommendation system, topic modeling, and sensor network localization. We propose a multilinear Bernoulli model, develop a rank-constrained likelihood-based estimation method, and obtain the theoretical accuracy guarantees. In contrast to continuous-valued problems, the binary tensor problem exhibits an interesting phase transition phenomenon according to the signal-to-noise ratio. The error bound for the parameter tensor estimation is established, and we show that the obtained rate is minimax optimal under the considered model. Furthermore, we develop an alternating optimization algorithm with convergence guarantees. The efficacy of our approach is demonstrated through both simulations and analyses of multiple data sets on the tasks of tensor completion and clustering.

[1]  F. L. Hitchcock The Expression of a Tensor or a Polyadic as a Sum of Products , 1927 .

[2]  R. Malach,et al.  When the Brain Loses Its Self: Prefrontal Inactivation during Sensorimotor Processing , 2006, Neuron.

[3]  Pauli Miettinen,et al.  Discovering facts with boolean tensor tucker decomposition , 2013, CIKM.

[4]  Xuan Bi,et al.  Multilayer tensor factorization with applications to recommender systems , 2017, The Annals of Statistics.

[5]  Yifan Zhang,et al.  Training Binary Weight Networks via Semi-Binary Decomposition , 2018, ECCV.

[6]  Andrea Montanari,et al.  A statistical model for tensor PCA , 2014, NIPS.

[7]  Hans-Peter Kriegel,et al.  A Three-Way Model for Collective Learning on Multi-Relational Data , 2011, ICML.

[8]  Hongtu Zhu,et al.  Tensor Regression with Applications in Neuroimaging Data Analysis , 2012, Journal of the American Statistical Association.

[9]  Tamara G. Kolda,et al.  Tensor Decompositions and Applications , 2009, SIAM Rev..

[10]  Anima Anandkumar,et al.  Tensor decompositions for learning latent variable models , 2012, J. Mach. Learn. Res..

[11]  Bo Huang,et al.  Square Deal: Lower Bounds and Improved Relaxations for Tensor Recovery , 2013, ICML.

[12]  Jan de Leeuw,et al.  Principal component analysis of binary data by iterated singular value decomposition , 2006, Comput. Stat. Data Anal..

[13]  L. Tucker,et al.  Some mathematical notes on three-mode factor analysis , 1966, Psychometrika.

[14]  Pauli Miettinen,et al.  Boolean Tensor Factorizations , 2011, 2011 IEEE 11th International Conference on Data Mining.

[15]  Khanh Dao Duc,et al.  OPERATOR NORM INEQUALITIES BETWEEN TENSOR UNFOLDINGS ON THE PARTITION LATTICE. , 2016, Linear algebra and its applications.

[16]  Wen-Xin Zhou,et al.  A max-norm constrained minimization approach to 1-bit matrix completion , 2013, J. Mach. Learn. Res..

[17]  David B. Dunson,et al.  Common and Individual Structure of Multiple Networks , 2017 .

[18]  Ewout van den Berg,et al.  1-Bit Matrix Completion , 2012, ArXiv.

[19]  Ming Yuan,et al.  Non-Convex Projected Gradient Descent for Generalized Low-Rank Tensor Regression , 2016, J. Mach. Learn. Res..

[20]  Kenneth Lange,et al.  Numerical analysis for statisticians , 1999 .

[21]  Ivan Oseledets,et al.  Tensor-Train Decomposition , 2011, SIAM J. Sci. Comput..

[22]  Tengyu Ma,et al.  On the optimization landscape of tensor decompositions , 2017, Mathematical Programming.

[23]  Anders M. Dale,et al.  An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest , 2006, NeuroImage.

[24]  M. McCarthy,et al.  Tensor decomposition for multi-tissue gene expression experiments , 2016, Nature Genetics.

[25]  Sanjoy Dasgupta,et al.  A Generalization of Principal Components Analysis to the Exponential Family , 2001, NIPS.

[26]  E. Candès,et al.  A modern maximum-likelihood theory for high-dimensional logistic regression , 2018, Proceedings of the National Academy of Sciences.

[27]  Jianhua Z. Huang,et al.  SPARSE LOGISTIC PRINCIPAL COMPONENTS ANALYSIS FOR BINARY DATA. , 2010, The annals of applied statistics.

[28]  Stephen P. Boyd,et al.  Convex Optimization , 2004, Algorithms and Theory of Computation Handbook.

[29]  David B. Dunson,et al.  Scalable Bayesian Low-Rank Decomposition of Incomplete Multiway Tensors , 2014, ICML.

[30]  Shmuel Friedland,et al.  Nuclear norm of higher-order tensors , 2014, Math. Comput..

[31]  M. Raichle,et al.  On the role of the corpus callosum in interhemispheric functional connectivity in humans , 2017, Proceedings of the National Academy of Sciences.

[32]  Yun S. Song,et al.  Three-way clustering of multi-tissue multi-individual gene expression data using constrained tensor decomposition , 2017, bioRxiv.

[33]  Geoffrey E. Hinton,et al.  Tensor Analyzers , 2013, ICML.

[34]  Trac D. Tran,et al.  Tensor sparsification via a bound on the spectral norm of random tensors , 2010, ArXiv.

[35]  Yun S. Song,et al.  Orthogonal Tensor Decompositions via Two-Mode Higher-Order SVD (HOSVD) , 2016, ArXiv.

[36]  Tengyu Ma,et al.  Matrix Completion has No Spurious Local Minimum , 2016, NIPS.

[37]  Han Liu,et al.  Provable sparse tensor decomposition , 2015, 1502.01425.

[38]  W. Denham,et al.  Multiple Measures of Alyawarra Kinship , 2005 .

[39]  Pauli Miettinen,et al.  Walk 'n' Merge: A Scalable Algorithm for Boolean Tensor Factorization , 2013, 2013 IEEE 13th International Conference on Data Mining.

[40]  Zhengwu Zhang,et al.  Common and individual structure of brain networks , 2017, The Annals of Applied Statistics.

[41]  Alexandre B. Tsybakov,et al.  Introduction to Nonparametric Estimation , 2008, Springer series in statistics.

[42]  David B. Dunson,et al.  Mapping population-based structural connectomes , 2018, NeuroImage.

[43]  Anru R. Zhang,et al.  Tensor SVD: Statistical and Computational Limits , 2017, IEEE Transactions on Information Theory.

[44]  P. McCullagh Regression Models for Ordinal Data , 1980 .

[45]  James C. Bezdek,et al.  Convergence of Alternating Optimization , 2003, Neural Parallel Sci. Comput..

[46]  Navid Ghadermarzy,et al.  Learning Tensors From Partial Binary Measurements , 2018, IEEE Transactions on Signal Processing.

[47]  Tamara G. Kolda,et al.  Scalable Tensor Factorizations with Missing Data , 2010, SDM.

[48]  Jean Lafond,et al.  Low Rank Matrix Completion with Exponential Family Noise , 2015, COLT.

[49]  Yong He,et al.  BrainNet Viewer: A Network Visualization Tool for Human Brain Connectomics , 2013, PloS one.

[50]  Fan Yang,et al.  Tensor and its tucker core: The invariance relationships , 2016, Numer. Linear Algebra Appl..

[51]  Adel Javanmard,et al.  1-bit matrix completion under exact low-rank constraint , 2015, 2015 49th Annual Conference on Information Sciences and Systems (CISS).

[52]  Christopher Yau,et al.  Probabilistic Boolean Tensor Decomposition , 2018, ICML.

[53]  Emmanuel Abbe,et al.  Community detection and stochastic block models: recent developments , 2017, Found. Trends Commun. Inf. Theory.

[54]  Aditya Bhaskara,et al.  Uniqueness of Tensor Decompositions with Applications to Polynomial Identifiability , 2013, COLT.

[55]  Chanwoo Lee,et al.  Tensor denoising and completion based on ordinal observations , 2020, ICML.

[56]  Tamara G. Kolda,et al.  Generalized Canonical Polyadic Tensor Decomposition , 2018, SIAM Rev..

[57]  Ryota Tomioka,et al.  Spectral norm of random tensors , 2014, 1407.1870.

[58]  A. Albert,et al.  On the existence of maximum likelihood estimates in logistic regression models , 1984 .

[59]  J. Kruskal Three-way arrays: rank and uniqueness of trilinear decompositions, with application to arithmetic complexity and statistics , 1977 .

[60]  Matthew Harding,et al.  Scalable Probabilistic Tensor Factorization for Binary and Count Data , 2015, IJCAI.

[61]  Hong Yan,et al.  Dimensionality reduction and topographic mapping of binary tensors , 2013, Pattern Analysis and Applications.

[62]  André Uschmajew,et al.  Local Convergence of the Alternating Least Squares Algorithm for Canonical Tensor Approximation , 2012, SIAM J. Matrix Anal. Appl..

[63]  Zenglin Xu,et al.  Distributed Flexible Nonlinear Tensor Factorization , 2016, NIPS.