Simple AE

class flood_forecast.meta_models.basic_ae.AE(input_shape: int, out_features: int)[source]
__init__(input_shape: int, out_features: int)[source]

A basic and simple to use AutoEncoder

Parameters:

  • input_shape (int) – The number of features for input.

  • out_features (int) – The number of output features (that will be merged)

forward(features: Tensor)[source]

Runs the full forward pass on the model. In practice this will only be done during training.

Parameters:

features ([type]) – [description]

Returns:

[description]

Return type:

[type]

generate_representation(features)[source]

A simple auto-encoder