Logistic regression exercise using Google Collab or similar tool

a. Import the tensorflow.keras libraries

b. Using tensorflow.keras.datasets import the MNIST dataset.

c. Plot in a 3x3 figure, random images of the MNIST dataset items. i.e.

192

785

459

d. Using sklearn.datasets.fetch_openml import the MNIST 784 dataset

e. Using the MNIST 784 dataset split the data in x_train, y_train, x_test, y_test and print your ratio of training and testing data.

f. Create a logistic regression model (sklearn) and train it using your train and test data splits.

g. Propose two metrics to evaluate the models and give a brief justification for these.

h. Calculate and print the two metrics.

i. Create a new logistic regression model and train it using 40% of the training data.

j. Calculate and print the two metrics.