Name
Abstract | ||
|---|---|---|
Most NDR algorithms need to solve large-scale eigenvalue problems or some variation of eigenvalue problems, which is of quadratic complexity of time and might be unpractical in case of large-size data sets. Besides, current algorithms are global, which are often sensitive to noise and disturbed by ill-conditioned matrix. In this paper, we propose a novel self-organizing NDR algorithm: SIE. The time complexity of SIE is O(NlogN). The main computing procedure of SIE is local, which improves the robustness of the algorithm remarkably. |
Year | DOI | Venue |
|---|---|---|
2005 | null | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
Keywords | Field | DocType |
main computing procedure,current algorithm,ill-conditioned matrix,quadratic complexity,large-scale eigenvalue problem,large-size data set,ndr algorithm,self-organizing approach,robust nonlinear dimension reduction,time complexity,eigenvalue problem,self organization,dimension reduction | Data set,Dimensionality reduction,Computer science,Matrix (mathematics),Self-organization,Robustness (computer science),Artificial intelligence,Time complexity,Eigenvalues and eigenvectors,Pattern recognition,Algorithm,Calculus,Geodesic | Conference |
Volume | Issue | ISSN |
3614 LNAI | null | 16113349 |
ISBN | Citations | PageRank |
3-540-28331-5 | 3 | 0.49 |
References | Authors | |
2 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Yuexian Hou | 1 | 269 | 38.59 |
| Liyue Yao | 2 | 3 | 0.49 |
| Pilian He | 3 | 29 | 7.46 |