Problem $1(1,0)$ : Build an MLP network with $8$ neurons in the first hidden layer, $4$ neurons each in the second and third hidden layers. Use $6$ neurons for the output layer. All the hidden layer neurons use ReLU activation function. The output layer neurons use softmax activation function. The network should be compiled using Adam optimizer, categorical$\_$crossentropy loss function and accuracy as a metric.

In

In $[]$:

def build$\_$nn$()$:

return:

model: A compiled keras model with the attributes provid

ed in the question

# Don't modify

model $=$ keras. Sequential $()$

## BEGIN SOLUTION

return model

Problem $2(1)$ Train the compiled network in problem $1$ in $10$ epochs, with a batch size of $32.$ Use the encoded training set for training $($utilize Helper function $2).$

In $[]$: def train$\_$model$($model$,$ encoded$\_$train$)$:

$,\text{'},$

parameters:

model: A compiled keras model with the attributes provid ed in the problem $2$

encoded$\_$train: The training dataset with $8$ columns. The first $6$ columns represent the one$-$hot$-$encoded region where the $p$ oint belongs.

The last $2$ columns represent the $(x,y)c$ o$-$ordinate of the point.

return:

model: A trained keras model with the provided settings in the question

X$\_$train: The encoded training $x$

y$\_$train: The encoded training $y$

### BEGIN SOLUTION

# Uncomment and complete the lines below

# X$\_$train $=$

#$y_{-}$train $=$

# Train the model below. Use X$\_$train, y$\_$train to train the $m$ odel with the given settings in the question.

# return model, x$\_$train, y$\_$train