Abstract
Visualization techniques for high-dimensional data sets play a pivotal role in exploratory analysis in a wide range of disciplines. A particularly challenging problem represents gene expression data based on microarray technology where the number of features (genes) typically exceeds 20,000, whereas the number of samples is frequently below 200. We investigated class-specific discrimination coefficients for each feature and each pair of classes for an effective nonlinear mapping to lower-dimensional space. We applied the technique to three microarray data sets and compared the projections to two-dimensional space with the results from a conventional multidimensional scaling method, a score plot resulting from principal component analysis, and projections from linear discriminant analysis. In the experiments, we observed that the discrimination coefficients allowed for an improved visualization of high-dimensional genomic data.
Original language | English |
---|---|
Pages (from-to) | 1211-1218 |
Number of pages | 8 |
Journal | Neural Computing and Applications |
Volume | 20 |
Issue number | 8 |
DOIs | |
Publication status | Published - 2011 Nov |
Externally published | Yes |
Fingerprint
Keywords
- Discrimination coefficients
- Linear discriminant analysis
- Microarrays
- Multidimensional scaling
- Principal component analysis
- Visualization
ASJC Scopus subject areas
- Artificial Intelligence
- Software
Cite this
Multidimensional scaling with discrimination coefficients for supervised visualization of high-dimensional data. / Berrar, Daniel; Ohmayer, Georg.
In: Neural Computing and Applications, Vol. 20, No. 8, 11.2011, p. 1211-1218.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Multidimensional scaling with discrimination coefficients for supervised visualization of high-dimensional data
AU - Berrar, Daniel
AU - Ohmayer, Georg
PY - 2011/11
Y1 - 2011/11
N2 - Visualization techniques for high-dimensional data sets play a pivotal role in exploratory analysis in a wide range of disciplines. A particularly challenging problem represents gene expression data based on microarray technology where the number of features (genes) typically exceeds 20,000, whereas the number of samples is frequently below 200. We investigated class-specific discrimination coefficients for each feature and each pair of classes for an effective nonlinear mapping to lower-dimensional space. We applied the technique to three microarray data sets and compared the projections to two-dimensional space with the results from a conventional multidimensional scaling method, a score plot resulting from principal component analysis, and projections from linear discriminant analysis. In the experiments, we observed that the discrimination coefficients allowed for an improved visualization of high-dimensional genomic data.
AB - Visualization techniques for high-dimensional data sets play a pivotal role in exploratory analysis in a wide range of disciplines. A particularly challenging problem represents gene expression data based on microarray technology where the number of features (genes) typically exceeds 20,000, whereas the number of samples is frequently below 200. We investigated class-specific discrimination coefficients for each feature and each pair of classes for an effective nonlinear mapping to lower-dimensional space. We applied the technique to three microarray data sets and compared the projections to two-dimensional space with the results from a conventional multidimensional scaling method, a score plot resulting from principal component analysis, and projections from linear discriminant analysis. In the experiments, we observed that the discrimination coefficients allowed for an improved visualization of high-dimensional genomic data.
KW - Discrimination coefficients
KW - Linear discriminant analysis
KW - Microarrays
KW - Multidimensional scaling
KW - Principal component analysis
KW - Visualization
UR - http://www.scopus.com/inward/record.url?scp=80054101306&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=80054101306&partnerID=8YFLogxK
U2 - 10.1007/s00521-010-0478-1
DO - 10.1007/s00521-010-0478-1
M3 - Article
AN - SCOPUS:80054101306
VL - 20
SP - 1211
EP - 1218
JO - Neural Computing and Applications
JF - Neural Computing and Applications
SN - 0941-0643
IS - 8
ER -