Name
Abstract | ||
|---|---|---|
•A KFDA kernel parameters optimization criterion is presented.•The uniformity of class-pair separabilities in kernel space is maximized.•The class separability in kernel space is maximized jointly.•Fourteen benchmark multiclass data sets are used for experiment.•The criterion can search the optimum KFDA kernel parameters more accurately. |
Year | DOI | Venue |
|---|---|---|
2013 | 10.1016/j.patrec.2013.03.005 | Pattern Recognition Letters |
Keywords | Field | DocType |
Kernel Fisher discriminant analysis (KFDA),Kernel parameter optimization,Feature extraction,Spectral regression kernel discriminant analysis (SRKDA) | Radial basis function kernel,Pattern recognition,Kernel embedding of distributions,Kernel Fisher discriminant analysis,Kernel principal component analysis,Polynomial kernel,Artificial intelligence,Kernel method,Variable kernel density estimation,Mathematics,Kernel (statistics) | Journal |
Volume | Issue | ISSN |
34 | 9 | 0167-8655 |
Citations | PageRank | References |
11 | 0.54 | 17 |
Authors | ||
3 | ||