Efficient parallel CP decomposition with pairwise perturbation and multi-sweep dimension tree
2021 IEEE International Parallel and Distributed Processing Symposium (IPDPS)(2021)
摘要
The widely used alternating least squares (ALS) algorithm for the canonical polyadic (CP) tensor decomposition is dominated in cost by the matricized-tensor times Khatri-Rao product (MTTKRP) kernel. This kernel is necessary to set up the quadratic optimization subproblems. State-of-the-art parallel ALS implementations use dimension trees to avoid redundant computations across MTTKRPs within each A...
更多查看译文
关键词
Tensors,Program processors,Perturbation methods,Approximation algorithms,Supercomputers,Computational efficiency,Parallel algorithms
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要