Kick-one-out-based variable selection method for Euclidean distance-based classifier in high-dimensional settings.
Tomoyuki NakagawaHiroki WatanabeMasashi HyodoPublished in: J. Multivar. Anal. (2021)
Keyphrases
- variable selection
- high dimensional
- cross validation
- dimension reduction
- pairwise
- prediction model
- feature space
- support vector machine
- feature selection
- clustering method
- unsupervised learning
- model selection
- feature set
- probabilistic model
- artificial neural networks
- training data
- similarity measure
- decision trees
- computer vision
- selection criterion