You wish to train a Perceptron on the given data using the Perceptron algorithm. Does the

algorithm converge? Answer without running the algorithm and justify your answer!

b$)$ Initialize the Perceptron weight and bias to $=0\mathrm{.}1$ and $=0\mathrm{.}1.$ Using step size $=0\mathrm{.}1,$

compute the values of the weight and bias after $1$ epoch of the Perceptron algorithm, iter$-$

ating through the inputs in the following order: $1,2,3,4.$

c$)$ Assume that an Adaline model trained to this data has weight and bias of $=0\mathrm{.}1$ and $=$

$0\mathrm{.}1.$ Compute the average loss over the training examples.

d$)$ Compare the average loss over the training examples before and after running one epoch

of gradient descent with step size $=0\mathrm{.}1.$ Initialize with the Adaline model from part $($c$)$