dataeval.detectors.ood.OOD_AE
=============================

.. py:class:: dataeval.detectors.ood.OOD_AE(model, device = None)



   Autoencoder based out-of-distribution detector.

   :param model: An Autoencoder model.
   :type model: Autoencoder


   .. py:method:: fit(x_ref, threshold_perc, loss_fn = None, optimizer = None, epochs = 20, batch_size = 64, verbose = False)

      Train the model and infer the threshold value.

      :param x_ref: Training data.
      :type x_ref: ArrayLike
      :param threshold_perc: Percentage of reference data that is normal.
      :type threshold_perc: float, default 100.0
      :param loss_fn: Loss function used for training.
      :type loss_fn: Callable | None, default None
      :param optimizer: Optimizer used for training.
      :type optimizer: Optimizer, default keras.optimizers.Adam
      :param epochs: Number of training epochs.
      :type epochs: int, default 20
      :param batch_size: Batch size used for training.
      :type batch_size: int, default 64
      :param verbose: Whether to print training progress.
      :type verbose: bool, default True