How to identify the amount of neurons and layers in a Neural Network

Question

I'm watching this lecture on youtube about neural networks. I'm unsure on how to identify the amount of neurons and layers in the network. I have posted the lecture slide from the youtube lecture on the left with 3 neurons and 2 layers. I found an image on google and posted it on the right. Are my guesses for the number neurons and layers correct for the google image?

Answer 1

Amount of neurons and layers are correct in your left (from lecture) image and on the second one you are wrong. There is three layers and twelve neurons, but whole number of neurons doesn't have any meaning, most important is how many neurons are in each layer. So you can describe this network like this:

Neural Network with three layers, first layer (input layer) consists with four neurons, second layer (hidden layer) - with six neurons and third layer (output layer) - with two neurons.

Even in this picture you have layers marked and well represented graphically

In any networks you have input layer and output layer, so you already have two layers and if inputs are being processed by another neurons before output neurons, it means that you have hidden layer. You can have many hidden layers in your net, it is called multi-layered neural net. Here is a examples some of them:

this is four layered neural network with two hidden layers

this is more complex network with three hidden layers. (Five layers in sum)

NOTE:

Increasing of layers in your network increases it's training time