Explain how many hidden neurons will be there in the network architecture. There will be $3$ input neurons and $4$ output neurons. Also please show the derivation$\mathrm{.}20$ POINT PROJECT

Create a neural network model of the rotary robot arm discussed in class using the state space model form.

Use the math model below as if it were the actual robot, to collect data from the robot

$\frac{d^2}{d{t}^2}=\frac{1}{J}T(t),(0)=0,(0)={}^{}(0)=0$

Integrated for const torques:

$(t)=\frac{T}{2J}{t}^2,$rad,$(t)=\frac{T}{J}t,ra\frac{d}{sec}$

$\frac{1}{2}$ page description of the network architecture $($# neurons, activations functions etc.$)$ and how

you arrived at it$,$ number of training pairs processed during training, training and testing RMS