Problem: MNIST dataset - The MNIST dataset is broken up into two parts - training, and test. Each part is made up of a series of images (28 x 28 pixel images of handwritten digits) and their respective labels (values from 0 - 9, representing which digit the image corresponds to).

In machine learning, we usually divide the training part to two sets of training and validation sets in order to adjust a machine learning model parameters. In your code, randomly select 20% of the training images along with their corresponding labels and name them as x_valid and y_valid, respectively. Name the remaining of the training images and their labels as x_train and y_train, respectively. Print the number of images in each training and validation set. Note: that there are no overlaps between the two sets. using:

from keras.datasets import mnist import matplotlib.pyplot as plt import matplotlib.gridspec as gridspec import numpy as np

(x_train, y_train), (x_test, y_test) = mnist.load_data()