Redesign topology builder

[Robbepop]

The current topology builder is not flexible enough to use it for building up complex neural networks involving different layer types such as Convolutonal Layers.

The new topology build infrastructure should enforce a mirroring between abstract topology layers and concrete neural network layers.

Some thoughts on this matter has been made, results so far are:

• There could be a need of some kind of InputLayer in topology structure to represent the initial layer.
• Fully connected layers could be renamed to DenseLayers which is shorter.
• There is a need for a generic ActivationLayer that is only responsible for serving the activation function. For identity activation functions this has the effect of simply leaving the activation layer away.
• At first DenseLayer and ActivationLayer are fit to replace the current system. However, this structure makes it possible to later add layer types for convolutional layer computation to the set such as 2-dimensional PoolingLayer and ConvolutionLayer.
• For interoperability between convolutional layers (2D) and normal layers (1D) there needs to be a kind of conversion between the 2-dimensional layers and the 1-dimensional layers.

[Robbepop]

An initial version of the new topology builder is now available.

It supports:

• FullyConnectedLayer: a.k.a. DenseLayer.
• ActivationLayer

It does not yet support the following layers:

• ContainerLayer: A good design for the topology API has yet to be found. Some designs are possible but there has not been a decision of preference, yet. Also ContainerLayer currently does not cover any practical use cases that aren't already possible.
• InputLayer: Input layers should not have an explicit notion in the topology builder. Besides that the current layer design does not require InputLayer at all.
• ConvolutionLayer: Not yet implemented, will require another design iteration.
• PoolingLayer: Not yet implemented.

For another design iteration a new issue will be opened.