Compara mètodes

	Regressió per passos	Regressió Lasso
Camp	Estadística	Aprenentatge automàtic
Família	Regression model	Machine learning
Any d'origen	1960	1996
Autor original	M. A. Efroymson	Tibshirani, R.
Tipus	Automated variable selection	Regularized linear regression (L1 penalty)
Font seminal	Efroymson, M. A. (1960). Multiple regression analysis. In A. Ralston & H. S. Wilf (Eds.), Mathematical Methods for Digital Computers (pp. 191–203). Wiley.	Tibshirani, R. (1996). Regression Shrinkage and Selection via the Lasso. Journal of the Royal Statistical Society: Series B, 58(1), 267–288.
Àlies	stepwise selection, forward stepwise regression, backward stepwise regression, forward-backward selection	LASSO Regresyonu, lasso, L1-regularized regression, L1 regularization
Relacionats	5	4
Resum	Stepwise regression is an automated variable selection procedure for multiple linear regression that adds or removes predictor variables one at a time according to a statistical criterion, typically the F-statistic or a p-value threshold. The forward-selection algorithm was formally described by Efroymson (1960) and the bidirectional variant was popularised by Draper and Smith in their landmark 1966 text Applied Regression Analysis. Despite widespread historical use, the method is now widely critiqued, making its documentation essential in any canonical methods library.	Lasso regression, introduced by Robert Tibshirani in 1996, is a linear regression method that adds an L1 penalty to the loss so that it shrinks coefficients and performs variable selection at the same time, producing a sparse model. By driving some coefficients exactly to zero it keeps only the predictors that matter.