A mixture of local and quadratic approximation variable selection algorithm in nonconcave penalized regression.
Assi N'GuessanIbrahim Sidi ZakariAbdallah MkhadriPublished in: ARIMA J. (2012)
Keyphrases
- selection algorithm
- selection scheme
- least squares
- resource selection
- model selection
- regression model
- linear functions
- linear regression
- search algorithm
- prototype selection
- loss function
- reduction method
- maximum likelihood
- mixture model
- pairwise
- regression problems
- approximation algorithms
- feature subset
- data sets
- feature subset selection
- data mining
- support vector
- objective function
- gaussian processes
- gaussian distribution
- fitness function
- active learning
- pattern recognition
- expected error
- multi variate