Solving a Mixture of Many Random Linear Equations by Tensor Decomposition and Alternating Minimization

Abstract

We consider the problem of solving mixed random linear equations with $k$ components. This is the noiseless setting of mixed linear regression. The goal is to estimate multiple linear models from mixed samples in the case where the labels (which sample corresponds to which model) are not observed. We give a tractable algorithm for the mixed linear equation problem, and show that under some technical conditions, our algorithm is guaranteed to solve the problem exactly with sample complexity linear in the dimension, and polynomial in $k$, the number of components. Previous approaches have required either exponential dependence on $k$, or super-linear dependence on the dimension. The proposed algorithm is a combination of tensor decomposition and alternating minimization. Our analysis involves proving that the initialization provided by the tensor method allows alternating minimization, which is equivalent to EM in our setting, to converge to the global optimum at a linear rate.

39 pages, 2 figures

Cited by in corpus (7)

• Principal component analysis for big data
• Max-Affine Regression: Provable, Tractable, and Near-Optimal Statistical Estimation
• Robust Meta-learning for Mixed Linear Regression with Small Batches
• Sample Complexity of Learning Mixtures of Sparse Linear Regressions
• Learning Mixtures of Linear Regressions in Subexponential Time via Fourier Moments
• Meta-learning for mixed linear regression
• Support Recovery of Sparse Signals from a Mixture of Linear Measurements