Deep Learning Toolbox class hierarchy

I am looking for some kind of condensed overview of the Deep Learning Toolbox class hierarchy. There are many types of network objects in the toolbox, and their descriptions are scattered across numerous pages in the online documentation. There seems to be a generic OG net object type such as the kind generated by feedforwardnet(). Then there are network types that trainNetwork() can generate such as SeriesNetworks and DAGNetworks from more recent generations of the toolbox. Even more recently there are dlnetworks, which seem to be the only type that can operate without an input or output layer. Below all these are more fundamental network building blocks like layers and layerGraphs.

I am looking for some kind of central document or video tutorial that discusses these objects, their different uses, and how they interrelate. Does such a thing exist?

Answer 1

[Muskan]

Hi Matt,

As per my understanding of the question, you can refer to the following documentations for a better understanding of the hierarchy in the Deep Learning Toolbox.

List of Deep Learning Layers - MATLAB & Simulink (mathworks.com)

Specify Layers of Convolutional Neural Network - MATLAB & Simulink (mathworks.com)

I hope the above information helps resolve your query.

Thanks

[Matt J]

Hi @Muskan,

No, these pages are all about the different layer object types, not the different network object types. They also don't talk about layerGraphs objects at all, as far as I can see.