Both the hierarchical and network models have their own classification.

Example: a neural network with too many layers may cause issues. Overfitting and how to prevent it are next.

Create UML diagram for this:  Layers of gasses are given, with certain type (ozone, oxygen, carbon dioxide) and thickness, affected by atmospheric variables (thunderstorm, sunshine, other effects). When a part of one layer changes into another layer due to an atmospheric variable, the newly transformed layer ascends and engrosses the first identical type of layer of gasses over it. In case there is no identical layer above, it creates a new layer on the top of the atmosphere. In the following we declare, how the different types of layers react to the different variables by changing their type and thickness. No layer can have a thickness less than 0.5 km, unless it ascends to the identical-type upper layer. In case there is no identical one, the layer perishes.

Can you name the three most important factors for a functional and productive network? Put it in your own words and show what you mean.