Least Absolute Shrinkage and Selection Operator

GPTKB entity

Predicate	Object
gptkbp:instanceOf	regression analysis method
gptkbp:abbreviation	gptkb:LASSO
gptkbp:application	feature selection high-dimensional data
gptkbp:category	supervised learning linear model
gptkbp:effect	shrinks some coefficients to zero
gptkbp:generalizes	gptkb:fused_LASSO gptkb:group_LASSO
https://www.w3.org/2000/01/rdf-schema#label	Least Absolute Shrinkage and Selection Operator
gptkbp:introduced	gptkb:Robert_Tibshirani 1996
gptkbp:limitation	can select at most n variables if n < p
gptkbp:method	gptkb:least_angle_regression coordinate descent
gptkbp:objective	minimize sum of squared errors plus L1 penalty
gptkbp:objective_function	least squares with L1 penalty
gptkbp:penalty_type	L1 norm
gptkbp:promotion	sparse solutions
gptkbp:relatedTo	gptkb:Elastic_Net gptkb:Ridge_regression
gptkbp:solvedBy	convex optimization
gptkbp:used_in	gptkb:machine_learning statistics
gptkbp:usedFor	regularization variable selection preventing overfitting
gptkbp:bfsParent	gptkb:Lasso_regression gptkb:Lasso_Regression gptkb:LASSO
gptkbp:bfsLayer	6