Loss, constraints and regularizers¶
These are some convenience functions that implement common losses, constraints and regularizers.
Smooth loss functions:
|
Logistic loss function. |
|
Squared loss. |
|
Huber loss |
Non-smooth terms accessed through their proximal operator
|
L1 norm, that is, the sum of absolute values: |
|
Group Lasso penalty |
|
Trace (aka nuclear) norm, sum of singular values |
|
Fused Lasso penalty |
|
2-dimensional Total Variation pseudo-norm |
Constraints can be incorporated in a similar way through
|
Indicator function over the L1 ball |
|
Projection onto the trace (aka nuclear) norm, sum of singular values |