2013 IEEE Symposium on Computational Intelligence and Data Mining (CIDM) 2013
DOI: 10.1109/cidm.2013.6597242
|View full text |Cite
|
Sign up to set email alerts
|

Interpretable magnification factors for topographic maps of high dimensional and structured data

Abstract: Visualisation of high-dimensional data is typically formulated as a non-linear mapping between a high-dimensional space and a two-dimensional latent space. The goal is that similar data items should be projected to similar coordinates in the latent space. Nevertheless, due to the non-linearity data items that are distant in the high-dimensional space may be projected close to each other in the latent space. Therefore, magnification factors must be analysed in order to detect stretches and contractions in the e… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
1
0

Year Published

2014
2014
2014
2014

Publication Types

Select...
2
1

Relationship

0
3

Authors

Journals

citations
Cited by 3 publications
(4 citation statements)
references
References 16 publications
0
1
0
Order By: Relevance
“…Examples of this may include Gaussian process latent variable models (GP-LVM, [16]) and dynamical models (GPDM, [17]) or temporal Laplacian eigenmaps ( [18]). It could also be extended to alternative display methods, such as the recently proposed cartograms 2 [19], [20] and topographic maps [21]. Following the line of this paper, we also aim to investigate the VB-GTM-TT model and its properties in more depth, focusing on its generalization capabilities and its use for prediction.…”
Section: Discussionmentioning
confidence: 99%
“…Examples of this may include Gaussian process latent variable models (GP-LVM, [16]) and dynamical models (GPDM, [17]) or temporal Laplacian eigenmaps ( [18]). It could also be extended to alternative display methods, such as the recently proposed cartograms 2 [19], [20] and topographic maps [21]. Following the line of this paper, we also aim to investigate the VB-GTM-TT model and its properties in more depth, focusing on its generalization capabilities and its use for prediction.…”
Section: Discussionmentioning
confidence: 99%
“…Examples of this may include GP-LVM [Lawrence, 2005]), GP dynamical models (GPDM, ) or temporal Laplacian eigenmaps [Lewandowski et al, 2010]. It could also be extended to alternative visual display methods, such as the Cartograms presented in chapter § 4, [García et al, 2013;Vellido, 2013, 2012] and warped topographic maps [Gianniotis, 2013].…”
Section: Discussionmentioning
confidence: 99%
“…One of the most interesting facts is that the distortion caused by the non-linear mapping can be explicitly quantified as a MF over the latent space used for data visualization. For this property, the concept of magnification has recently been applied to manifold learning methods for NLDR, in order to visualize the distortion due to the embedding of a manifold in a high-dimensional space [Tosi and Vellido, 2012;Tosi and Vellido, 2013;Gianniotis, 2013;Tosi et al, 2014b,a]. More details about novel visualization techniques involving MF are presented in the following chapter.…”
Section: Magnification Factorsmentioning
confidence: 99%
“…Other possibilities might well be considered. Some research in that direction has only recently been published [92].…”
Section: Suggestions For Future Researchmentioning
confidence: 99%