Doubly Decomposing Nonparametric Tensor Regression
arXiv:1506.05967
Abstract
Nonparametric extension of tensor regression is proposed. Nonlinearity in a high-dimensional tensor space is broken into simple local functions by incorporating low-rank tensor decomposition. Compared to naive nonparametric approaches, our formulation considerably improves the convergence rate of estimation while maintaining consistency with the same function class under specific conditions. To estimate local functions, we develop a Bayesian estimator with the Gaussian process prior. Experimental results show its theoretical properties and high performance in terms of predicting a summary statistic of a real complex network.
arXiv comment21 pages
References in corpus (5)
- Tensor Regression with Applications in Neuroimaging Data Analysis
- Rates of contraction of posterior distributions based on Gaussian process priors
- Sparsity in multiple kernel learning
- Fast learning rate of multiple kernel learning: Trade-off between sparsity and smoothness
- Learning Tensors in Reproducing Kernel Hilbert Spaces with Multilinear Spectral Penalties
Cited by in corpus (9)
- Tensor Networks for Dimensionality Reduction and Large-Scale Optimizations. Part 2 Applications and Future Perspectives
- Broadcasted Nonparametric Tensor Regression
- CP Degeneracy in Tensor Regression
- Bayesian Nonlinear Tensor Regression with Functional Fused Elastic Net Prior
- Bayesian Methods in Tensor Analysis
- Linear Tensor Projection Revealing Nonlinearity
- Fast and Scalable Estimator for Sparse and Unit-Rank Higher-Order Regression Models
- Sparse and Low-Rank High-Order Tensor Regression via Parallel Proximal Method
- Soft Tensor Regression