Using α = 1 corresponds to pure L 1 regularization, and α = 0 corresponds to pure L 2 regularization. Any value in-between is then an Elastic Net regularization. @param alpha the value in [0, 1] for determining the regularization penalty's interpolation between pure
(double alpha)
| 196 | * regularization. |
| 197 | */ |
| 198 | public void setAlpha(double alpha) |
| 199 | { |
| 200 | if(alpha < 0 || alpha > 1 || Double.isNaN(alpha)) |
| 201 | throw new IllegalArgumentException("alpha must be in [0, 1], not " + alpha); |
| 202 | |
| 203 | this.alpha = alpha; |
| 204 | } |
| 205 | |
| 206 | /*** |
| 207 | * |
no outgoing calls