yamle.methods.delta_uq module

class yamle.methods.delta_uq.DeltaUQMethod(**kwargs)

Bases: MCSamplingMethod

This class is the extension of the base method for delta-UQ method.

The core of the method is in applying anchors during training and inference, the anchors are the samples themselves where the input of the network is: [x-anchor, anchor] where anchor is reshufled x.

Note that, this method requires a special treatment in case of data augmentation. The anchors during test should be from a training set and not from the test set. Hence we will cache some of the training samples.

state_dict()

This method is used to get the state of the method.

Return type:

Dict[str, Any]

load_state_dict(state_dict)

This method is used to load the state of the method.

Return type:

None

static add_specific_args(parent_parser)

This method is used to add the specific arguments for the class.

Return type:

ArgumentParser