Name
Abstract | ||
|---|---|---|
Novel parallel algorithms for tensor completion problems, with applications to recommender systems and function learning.Parallelization strategy offers greatly reduced memory requirements compared to previously published matrix equivalents.Convergence results for both alternating least squares and cyclic coordinate descent. Low-rank tensor completion addresses the task of filling in missing entries in multi-dimensional data. It has proven its versatility in numerous applications, including context-aware recommender systems and multivariate function learning. To handle large-scale datasets and applications that feature high dimensions, the development of distributed algorithms is central. In this work, we propose novel, highly scalable algorithms based on a combination of the canonical polyadic (CP) tensor format with block coordinate descent methods. Although similar algorithms have been proposed for the matrix case, the case of higher dimensions gives rise to a number of new challenges and requires a different paradigm for data distribution. The convergence of our algorithms is analyzed and numerical experiments illustrate their performance on distributed-memory architectures for tensors from a range of different applications. |
Year | DOI | Venue |
|---|---|---|
2016 | 10.1016/j.parco.2015.10.002 | Parallel Computing |
Keywords | Field | DocType |
Low-rank tensor completion,Canonical tensor format,Parallel tensor completion,Parallel cyclic coordinate descent,Parallel alternating least squares | Convergence (routing),Recommender system,Tensor,Parallel algorithm,Computer science,Matrix (mathematics),Filling-in,Algorithm,Theoretical computer science,Distributed algorithm,Coordinate descent | Journal |
Volume | Issue | ISSN |
57 | C | 0167-8191 |
Citations | PageRank | References |
13 | 0.53 | 16 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Lars Karlsson | 1 | 51 | 5.16 |
| Daniel Kressner | 2 | 449 | 48.01 |
| André Uschmajew | 3 | 135 | 9.34 |