deepod.models.PReNetTS

class deepod.models.PReNetTS(epochs=100, batch_size=64, lr=0.001, network='Transformer', seq_len=30, stride=1, rep_dim=128, hidden_dims='512', act='GELU', bias=False, n_heads=8, d_model=512, attn='self_attn', pos_encoding='fixed', norm='BatchNorm', epoch_steps=-1, prt_steps=10, device='cuda', verbose=2, random_state=42)[source]

Deep Weakly-supervised Anomaly Detection (KDD‘23)

Parameters:

  • epochs (int) – The number of epochs for training the model. Default is 100.

  • batch_size (int) – The size of the batch for training. Default is 64.

  • lr (float) – The learning rate. Default is 1e-3.

  • network (str) – The type of network used, ‘Transformer’ by default.

  • seq_len (int) – The length of the input sequences. Default is 30.

  • stride (int) – The stride for sliding window on data. Default is 1.

  • rep_dim (int) – The representation dimension. Default is 128.

  • hidden_dims (str) – The hidden layer dimensions, separated by commas. Default is ‘512’.

  • act (str) – The activation function. Default is ‘GELU’.

  • bias (bool) – Whether to use bias in the layers. Default is False.

  • n_heads (int) – The number of attention heads in a transformer. Default is 8.

  • d_model (int) – The dimensionality of the transformer model. Default is 512.

  • attn (str) – The type of attention mechanism. Default is ‘self_attn’.

  • pos_encoding (str) – The type of position encoding. Default is ‘fixed’.

  • norm (str) – The type of normalization layer. Default is ‘BatchNorm’.

  • epoch_steps (int) – The steps per epoch, -1 indicates using the full dataset. Default is -1.

  • prt_steps (int) – The steps for printing during training. Default is 10.

  • device (str) – The device for training, ‘cuda’ by default.

  • verbose (int) – The verbosity level. Default is 2.

  • random_state (int) – The random state seeding. Default is 42.

Methods

__init__([epochs, batch_size, lr, network, ...])

Initializes the PReNetTS model with specified parameters.

decision_function(X[, return_rep])

Predict raw anomaly scores of X using the fitted detector.

decision_function_update(z, scores)

for any updating operation after decision function

epoch_update()

for any updating operation after each training epoch

fit(X[, y])

Fit detector.

fit_auto_hyper(X[, y, X_test, y_test, ...])

Fit detector.

inference_forward(batch_x, net, criterion)

Processes a single inference batch through the model.

inference_prepare(X)

Prepares the model for inference by setting up the test data loader.

load_model(path)

load_ray_checkpoint(best_config, best_checkpoint)

predict(X[, return_confidence])

Predict if a particular sample is an outlier or not.

save_model(path)

set_seed(seed)

set_tuned_net(config)

set_tuned_params()

training_forward(batch_x, net, criterion)

Processes a single training batch through the model.

training_prepare(X, y)

Prepares the model for training by setting up the data loader, network, and criterion.
decision_function(X, return_rep=False)

Predict raw anomaly scores of X using the fitted detector.

The anomaly score of an input sample is computed based on the fitted detector. For consistency, outliers are assigned with higher anomaly scores.

Parameters:

  • X (numpy array of shape (n_samples, n_features)) – The input samples. Sparse matrices are accepted only if they are supported by the base estimator.

  • return_rep (boolean, optional, default=False) – whether return representations

Returns:

anomaly_scores – The anomaly score of the input samples.

Return type:

numpy array of shape (n_samples,)

decision_function_update(z, scores)

for any updating operation after decision function

epoch_update()

for any updating operation after each training epoch

fit(X, y=None)

Fit detector. y is ignored in unsupervised methods.

Parameters:

  • X (numpy array of shape (n_samples, n_features)) – The input samples.

  • y (numpy array of shape (n_samples, )) – Not used in unsupervised methods, present for API consistency by convention. used in (semi-/weakly-) supervised methods

Returns:

self – Fitted estimator.

Return type:

object

fit_auto_hyper(X, y=None, X_test=None, y_test=None, n_ray_samples=5, time_budget_s=None)

Fit detector. y is ignored in unsupervised methods.

Parameters:

  • X (numpy array of shape (n_samples, n_features)) – The input samples.

  • y (numpy array of shape (n_samples, )) – Not used in unsupervised methods, present for API consistency by convention. used in (semi-/weakly-) supervised methods

  • X_test (numpy array of shape (n_samples, n_features), default=None) – The input testing samples for hyper-parameter tuning.

  • y_test (numpy array of shape (n_samples, ), default=None) – Label of input testing samples for hyper-parameter tuning.

  • n_ray_samples (int, default=5) – Number of times to sample from the hyperparameter space

  • time_budget_s (int, default=None) – Global time budget in seconds after which all trials of Ray are stopped.

Returns:

config – tuned hyper-parameter

Return type:

dict

inference_forward(batch_x, net, criterion)[source]

Processes a single inference batch through the model.

Parameters:

  • batch_x (tuple) – A tuple of the batch data.

  • net (torch.nn.Module) – The network model.

  • criterion (callable) – The loss criterion used for evaluation.

Returns:

A tuple containing the batch data and the computed scores.

Return type:

tuple

inference_prepare(X)[source]

Prepares the model for inference by setting up the test data loader.

Parameters:

X (numpy.ndarray) – Test data.

Returns:

A list of batches for testing.

Return type:

list

predict(X, return_confidence=False)

Predict if a particular sample is an outlier or not.

Parameters:

  • X (numpy array of shape (n_samples, n_features)) – The input samples.

  • return_confidence (boolean, optional(default=False)) – If True, also return the confidence of prediction.

Returns:

  • outlier_labels (numpy array of shape (n_samples,)) – For each observation, tells whether it should be considered as an outlier according to the fitted model. 0 stands for inliers and 1 for outliers.

  • confidence (numpy array of shape (n_samples,).) – Only if return_confidence is set to True.

training_forward(batch_x, net, criterion)[source]

Processes a single training batch through the model.

Parameters:

  • batch_x (tuple) – A tuple of the batch data.

  • net (torch.nn.Module) – The network model.

  • criterion (callable) – The loss criterion.

Returns:

The computed loss for the batch.

Return type:

Tensor

training_prepare(X, y)[source]

Prepares the model for training by setting up the data loader, network, and criterion.

Parameters:

  • X (numpy.ndarray) – Training data.

  • y (numpy.ndarray) – Training labels.

Returns:

A tuple containing the training data loader, the network model, and the loss criterion.

Return type:

tuple