Name
Abstract | ||
|---|---|---|
As a popular heuristic to the matrix rank minimization problem, nuclear norm minimization attracts intensive research attentions. Matrix factorization based algorithms can reduce the expensive computation cost of SVD for nuclear norm minimization. However, most matrix factorization based algorithms fail to provide the theoretical guarantee for convergence caused by their non-unique factorizations. This paper proposes an efficient and accurate Linearized Grassmannian Optimization (Lingo) algorithm, which adopts matrix factorization and Grassmann manifold structure to alternatively minimize the subproblems. More specially, linearization strategy makes the auxiliary variables unnecessary and guarantees the close-form solution for low per-iteration complexity. Lingo then converts linearized objective function into a nuclear norm minimization over Grassmannian manifold, which could remedy the non-unique of solution for the low-rank matrix factorization. Extensive comparison experiments demonstrate the accuracy and efficiency of Lingo algorithm. The global convergence of Lingo is guaranteed with theoretical proof, which also verifies the effectiveness of Lingo. |
Year | DOI | Venue |
|---|---|---|
2015 | 10.1145/2806416.2806532 | ACM International Conference on Information and Knowledge Management |
Field | DocType | Citations |
Convergence (routing),Singular value decomposition,Mathematical optimization,Heuristic,Matrix completion,Computer science,Matrix decomposition,Matrix norm,Grassmannian,Linearization | Conference | 5 |
PageRank | References | Authors |
0.42 | 11 | 6 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Qian Li | 1 | 32 | 8.29 |
| Wenjia Niu | 2 | 178 | 30.33 |
| Gang Li | 3 | 381 | 62.77 |
| Ya-nan Cao | 4 | 131 | 19.42 |
| Jianlong Tan | 5 | 132 | 22.14 |
| Li Guo | 6 | 102 | 10.23 |