Neural Network Definition

class omlt.neuralnet.network_definition.NetworkDefinition(scaling_object=None, scaled_input_bounds=None, unscaled_input_bounds=None)

Bases: object

add_edge(from_layer, to_layer)

Add an edge between two layers.

Parameters:

• from_layer (Layer) – the layer with the outbound connection

• to_layer (Layer) – the layer with the inbound connection

add_layer(layer)

Add a layer to the network.

Parameters:

layer (Layer) – the layer to add to the network

property input_layers

Return an iterator over the input layers.

property input_nodes

An alias for input_layers.

layer(layer_id)

Return the layer with the given id.

property layers

Return an iterator over all the layers.

property output_layers

Return an iterator over the output layer.

property output_nodes

An alias for output_layers.

predecessors(layer)

Predecessors.

Return an iterator over the layers with outbound connections into the layer.

property scaled_input_bounds

Scaled Input Bounds.

Return a dict of tuples containing lower and upper bounds of neural network inputs.

property scaling_object

Return an instance of the scaling object supporting the ScalingInterface.

successors(layer)

Successors.

Return an iterator over the layers with an inbound connection from the layer.

property unscaled_input_bounds

Unscaled Input Bounds.

Return a dict of tuples containing lower and upper bounds of unscaled neural network inputs.