Name
Abstract | ||
|---|---|---|
Initialization of self-organizing maps is typically based on random vectors within the given input space. The implicit problem with random initialization is the overlap (entanglement) of connections between neurons. In this paper, we present a new method of initialization based on a set of self-similar curves known as Hilbert curves. Hilbert curves can be scaled in network size for the number of neurons based on a simple recursive (fractal) technique, implicit in the properties of Hilbert curves. We have shown that when using Hilbert curve vector (HCV) initialization in both classical SOM algorithm and in a parallel-growing algorithm (ParaSOM), the neural network reaches better coverage and faster organization. |
Year | DOI | Venue |
|---|---|---|
2010 | 10.1007/s00521-010-0413-5 | Neural Computing and Applications |
Keywords | DocType | Volume |
parallel-growing algorithm,neural network,hilbert curvesself-organizing maps � initializationneural networks,Fractal initialization,high-quality mapping,Hilbert curve,better coverage,random initialization,classical SOM algorithm,network size,self-organizing map,Hilbert curve vector,implicit problem,random vector | Journal | 19 |
Issue | ISSN | Citations |
7 | 1433-3058 | 7 |
PageRank | References | Authors |
0.61 | 18 | 4 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Iren Valova | 1 | 136 | 25.44 |
| Derek Beaton | 2 | 47 | 5.52 |
| Alexandre Buer | 3 | 28 | 2.04 |
| Daniel MacLean | 4 | 113 | 11.82 |