You are tasked to build an image classifier for the MNIST dataset of handwritten numbers, implementing thek-nearest neighbors (k-NN)algorithm. You will need the following:

• The MNIST dataset, available on multiple servers on the Internet. For example:
• http://yann.lecun.com/exdb/mnist/
• http://www.pymvpa.org/datadb/mnist.html
• The Python packageneighbors.KNeighborsClassifier:https://scikit-learn.org/stable/modules/generated/sklearn.neighbors.KNeighborsClassifier.html

The input to your classifier program is an image containing a digit, 0-9. Your program must correctly identify the digit with an accuracy of 95%. Here the outline of your task, but you will have to do a bit of research on your own (and increasingly so throughout the program) to fill in the details:

Familiarize yourself with the MNIST dataset

Familiarize yourself with the k-NN algorithm and its Python implementation in sklearn

USE PYTHON and implement the k-NN algorithm:

• Import the package kNeighborsClassifier.
• Be mindful of the train-test split and set the parameters accordingly (justify your choice).
• Identify the variables in the dataset and define the Euclidean distance between an element in the test set and the training set.
• Calculate the distance between the test element and each of if its k nearest neighbors.
• Count the occurrence of each digit within the k nearest neighbors and identify the most popular digit.
• Identify the test element as the digit voted as most popular in the set of the k nearest neighbors.
• Classify the test element accordingly (i.e. based on the popular vote).
• Calculate the error.