Question $1[1$ pt$]$ Figure $1$ shows a sample dataset with three types of instances $($rectangles$,$

diamonds, and circles$).$ The instances are also color coded $($each color denotes one type of

instances$).$ Each instance has two feature dimensions $($X$1$ and X$2)$ as shown in Figure $1.$ Design a

one hidden layer feedforward neural network to correctly classify instances into groups $($Only

design the network structure. No need to find correct weight values$)$

$$ Show your network $($including input layer, hidden layer, and output layers$).$ Please show

correct input dimensions and number of nodes of each layer $[0\mathrm{.}5$ pt$]$

$$ Explain the functionality of each hidden layer. Explain why do you choose a certain

number of hidden nodes and output nodes? $[0\mathrm{.}5$ pt$]$