nexuml_library.scenarios.asd.synthetic_conv_ae

nexuml_library.scenarios.asd.synthetic_conv_ae

Synthetic convolutional autoencoder scenarios.

synthetic_conv_ae_anomaly

synthetic_conv_ae_anomaly(
    input_shape: tuple[int, int, int] = (1, 64, 64),
    num_samples: int = 160,
    anomaly_fraction: float = 0.2,
    latent_dim: int = 64,
    channel_schedule: list[int] | None = None,
    lr: float = 0.001,
    batch_size: int = 16,
    max_epochs: int = 2,
) -> ScenarioSpec

Synthetic spectrogram convolutional AE anomaly workflow.

Returns:

Name	Type	Description
ScenarioSpec	ScenarioSpec	Assembled scenario with pipeline, training, data and evaluation.

synthetic_conv_cvae_anomaly

synthetic_conv_cvae_anomaly(
    input_shape: tuple[int, int, int] = (1, 64, 64),
    num_samples: int = 160,
    anomaly_fraction: float = 0.2,
    encoder_dim: int = 128,
    latent_dim: int = 32,
    beta: float = 0.001,
    channel_schedule: list[int] | None = None,
    lr: float = 0.001,
    batch_size: int = 16,
    max_epochs: int = 2,
) -> ScenarioSpec

Synthetic spectrogram convolutional VAE anomaly workflow.

Returns:

Name	Type	Description
ScenarioSpec	ScenarioSpec	Assembled scenario with pipeline, training, data and evaluation.