regression
Lasso regression
Fit a linear regression with coefficient shrinkage and possible variable selection.
Requirements
- continuous outcome
- predictor(s)
- penalty alpha
Avoid when
- highly correlated predictors where ridge is preferred
- unpenalized inference with p-values is required
Formula
arg minβ ||y−Xβ||² + λ||β||₁
Interpretation
Some coefficients may be shrunk exactly to zero; p-values are not reported.
Worked example
Fit a sparse continuous-outcome model when predictor selection is exploratory.
Open example datasets