elasticnet

  • ==Combines both of lasso and ridge==
  • It will try to learn a sparse model like Lasso and retain regularization properties of ridge
  • Tuned with $\alpha$ and $l1_{ratio}$ parameters