accelbrainbase.observabledata._mxnet.adversarialmodel.discriminativemodel package¶
Submodules¶
accelbrainbase.observabledata._mxnet.adversarialmodel.discriminativemodel.eb_discriminative_model module¶
-
class
accelbrainbase.observabledata._mxnet.adversarialmodel.discriminativemodel.eb_discriminative_model.EBDiscriminativeModel(model, initializer=None, learning_rate=1e-05, optimizer_name='SGD', hybridize_flag=True, scale=1.0, ctx=gpu(0), **kwargs)¶ Bases:
accelbrainbase.observabledata._mxnet.adversarialmodel.discriminative_model.DiscriminativeModelDiscriminative model, which discriminates true from fake, in the Energy-based Generative Adversarial Network(EBGAN).
The Energy-based Generative Adversarial Network (EBGAN) model(Zhao, J., et al., 2016) which views the discriminator as an energy function that attributes low energies to the regions near the data manifold and higher energies to other regions. The Auto-Encoders have traditionally been used to represent energy-based models. When trained with some regularization terms, the Auto-Encoders have the ability to learn an energy manifold without supervision or negative examples. This means that even when an energy-based Auto-Encoding model is trained to reconstruct a real sample, the model contributes to discovering the data manifold by itself.
References
- Zhao, J., Mathieu, M., & LeCun, Y. (2016). Energy-based generative adversarial network. arXiv preprint arXiv:1609.03126.
-
forward_propagation(F, x)¶ Hybrid forward with Gluon API.
Parameters: - F – mxnet.ndarray or mxnet.symbol.
- x – mxnet.ndarray of observed data points.
Returns: mxnet.ndarray or mxnet.symbol of inferenced feature points.
-
hybrid_forward(F, x)¶ Hybrid forward with Gluon API.
Parameters: - F – mxnet.ndarray or mxnet.symbol.
- x – mxnet.ndarray of observed data points.
Returns: mxnet.ndarray or mxnet.symbol of inferenced feature points.