ElasticNet

Combines both L1 regularisation and L2 regularisation

The formula contains the Mean Square Error, L1 term and L2 term.

• MSE: $\sum _{i=1}^n(y_i-y(x_i){)}^2$ for basic error term
• L1 term: $\alpha \sum _{j=1}^p|w_j|$ for feature selection
• L2 term: $\alpha \sum _{j=1}^p(w_j{)}^2$ for weight shrinkage