Accelerating Block Coordinate Descent for Nonnegative Tensor Factorization

Abstract

This paper is concerned with improving the empirical convergence speed of block-coordinate descent algorithms for approximate nonnegative tensor factorization (NTF). We propose an extrapolation strategy in-between block updates, referred to as heuristic extrapolation with restarts (HER). HER significantly accelerates the empirical convergence speed of most existing block-coordinate algorithms for dense NTF, in particular for challenging computational scenarios, while requiring a negligible additional computational budget.

32 pages, 24 figures

References in corpus (4)

• Nonnegative Matrix Factorization for Signal and Data Analytics: Identifiability, Algorithms, and Applications
• Accelerating Nonnegative Matrix Factorization Algorithms using Extrapolation
• Pencil-based algorithms for tensor rank decomposition are not stable
• Accelerating Block Coordinate Descent for Nonnegative Tensor Factorization

Cited by in corpus (4)

• Computing Large-Scale Matrix and Tensor Decomposition with Structured Factors: A Unified Nonconvex Optimization Perspective
• Accelerating Block Coordinate Descent for Nonnegative Tensor Factorization
• Bounded Simplex-Structured Matrix Factorization: Algorithms, Identifiability and Applications
• Representation Theorem for Matrix Product States