A novel method for feature selection with random sampling HDMR and its application to hyperspectral image classification.
Gülsen Taskin KayaHüseyin KayaLorenzo BruzzonePublished in: IGARSS (2015)
Keyphrases
- random sampling
- hyperspectral image classification
- method for feature selection
- active learning
- feature selection
- sampling algorithm
- semi supervised
- sampling procedure
- hyperspectral images
- random sample
- hyperspectral
- sampling methods
- random samples
- feature subset
- random projections
- sample size
- machine learning
- learning algorithm
- high dimensional
- lower bound
- pattern recognition
- stratified sampling
- data mining
- support vector machine
- training set