Solve the following set of problems using Python and submit the code file

with extension $.$ipynb.

$1.$ Load the MNIST Digit dataset, show the size of the training and test set.

$2.$ Develop a one hidden layer multi$-$layer perceptron model on the above training data,

report the accuracy of the model.

$3.$ From question $2,$ set the number of hidden layers of the MLP model as $[2,4,6,8,10],$

set the hidden layer size as $100,$ show the accuracies on the test set.

$4.$ From question $2,$ for the hidden layer set the hidden layer size as $[50,100,150,200],$

show the accuracies on the test set.

$5.$ Based on question $3$ and $4,$ explain the key findings.