MNIST Dataset and Python Assignment

Assignment Task

Download the MNIST data set (both training and test sets and labels).

The database contains 70,000 28x28 black and white images representing the digits zero through nine. The data is split into two subsets, with 60,000 images belonging to the training set and 10,000 images belonging to the testing set. As it can be seen from the image above, the handwritten digits consist of varying styles and complexities.

The labels will tell you which number it is: 1,2, 9, 0. Let each output be denoted by the vector.

Let B be the set of output vectors.

B = [y1 y2 …yn]

And let the matrix A be the corresponding reshaped (vectorized) MNIST images

A = [x1 x2 …. xn]

Thus each vector xj is a vector reshaped from the nxn MNIST image.

1. Using various AX=B solvers, determine a mapping from the image space to the label space.

2. By promoting sparsity, determine and rank which pixels in the MNIST set are most informative for correctly labeling the numbers (You may come up with your own heuristics or empirical rules for this. Use MATLAB command color to visualize the results from X).

3. Apply the most important pixels to the test data set to see how accurate your are with as few pixels as possible.

4. Redo the analysis with each number individually to find the most important pixels for each number.

5. Think about the interpretation of what you can do with is AX = B problem.

This IT and Computer Science has been solved by our PHD Experts at My Uni Paper.