supervised

The supervised approach trains solely on the labeled source domain. It can be used for pre-training or as a baseline to compare adaption approaches against.

Data --> FeatureExtractor --> Features --> Regressor  --> RUL Prediction

SupervisedApproach

Bases: AdaptionApproach

The supervised approach uses either MSE, MAE or RMSE loss to train a feature extractor and regressor in a supervised fashion on the source domain. It can be used either for pre-training or as a baseline to compare adaption approaches against.

The regressor needs the same number of input units as the feature extractor has output units.

Examples:

>>> from rul_adapt import model
>>> from rul_adapt import approach
>>> feat_ex = model.CnnExtractor(1, [16, 16, 1], 10, fc_units=16)
>>> reg = model.FullyConnectedHead(16, [1])
>>> disc = model.FullyConnectedHead(16, [8, 1], act_func_on_last_layer=False)
>>> main = approach.SupervisedApproach("mse")
>>> main.set_model(feat_ex, reg, disc)

__init__(loss_type, rul_scale=1, rul_score_mode='phm08', evaluate_degraded_only=False, **optim_kwargs)

Create a supervised approach.

The regressor output can be scaled with rul_scale to control its magnitude. By default, the RUL values are not scaled.

For more information about the possible optimizer keyword arguments, see here.

Parameters:

Name	Type	Description	Default
rul_score_mode	Literal['phm08', 'phm12']		'phm08'
loss_type	Literal['mse', 'mae', 'rmse']	Training loss function to use. Either 'mse', 'mae' or 'rmse'.	required
rul_scale	int	Scalar to multiply the RUL prediction with.	1
evaluate_degraded_only	bool	Whether to only evaluate the RUL score on degraded samples.	False
**optim_kwargs	Any	Keyword arguments for the optimizer, e.g. learning rate.	{}

test_step(batch, batch_idx, dataloader_idx=-1)

Execute one test step.

The batch argument is a list of two tensors representing features and labels. A RUL prediction is made from the features and the validation RMSE and RUL score are calculated. The metrics recorded for dataloader idx zero are assumed to be from the source domain and for dataloader idx one from the target domain. The metrics are written to the configured logger under the prefix test.

Parameters:

Name	Type	Description	Default
batch	List[Tensor]	A list containing a feature and a label tensor.	required
batch_idx	int	The index of the current batch.	required
dataloader_idx	int	The index of the current dataloader (0: source, 1: target).	-1

training_step(batch, batch_idx)

Execute one training step.

The batch argument is a list of two tensors representing features and labels. The features are used to predict RUL values that are compared against the labels with the specified training loss. The loss is then logged.

Parameters:

Name	Type	Description	Default
batch	List[Tensor]	A list of feature and label tensors.	required
batch_idx	int	The index of the current batch.	required

Returns: The training loss.

validation_step(batch, batch_idx, dataloader_idx=-1)

Execute one validation step.

The batch argument is a list of two tensors representing features and labels. The features are used to predict RUL values that are compared against the labels with an RMSE loss. The loss is then logged.

Parameters:

Name	Type	Description	Default
batch	List[Tensor]	A list of feature and label tensors.	required
batch_idx	int	The index of the current batch.	required