Automatic selection of parameters in LLE
Locally Linear Embedding (LLE) is a nonlinear dimensionality reduction technique, which preserves the local geometry of high dimensional space performing an embedding to low dimensional space. LLE algorithm has 3 free parameters that must be set to calculate the embedding: the number of nearest neig...
Saved in:
Published in | Revista Facultad de Ingeniería no. 56; pp. 170 - 181 |
---|---|
Main Authors | , , , , |
Format | Journal Article |
Language | English |
Published |
Universidad de Antioquia
28.02.2013
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Locally Linear Embedding (LLE) is a nonlinear dimensionality reduction technique, which preserves the local geometry of high dimensional space performing an embedding to low dimensional space. LLE algorithm has 3 free parameters that must be set to calculate the embedding: the number of nearest neighbors k, the output space dimensionality m and the regularization parameter a. The last one only is necessary when the value of k is greater than the dimensionality of input space or data are not located in general position, and it plays an important role in the embedding results. In this paper we propose a pair of criteria to find the optimum value for the parameters kand a, to obtain an embedding that faithfully represent the input data space. Our approaches are tested on 2 artificial data sets and 2 real world data sets to verify the effectiveness of the proposed criteria, besides the results are compared against methods found in the state of art. |
---|---|
ISSN: | 0120-6230 2422-2844 |
DOI: | 10.17533/udea.redin.14665 |