I need a MATLAB code. So the program that needs to be used is MATLAB

Design and implement a Hopfield network that memorizes digits and alphabets and retrieves them. Each digit or alphabet will be represented as a 10 x 12 matrix containing 0s or 1s. The network will recover the digit or alphabet when presented with the image with a few bits flipped.

1.Create the patterns of the digits/alphabets and memorize them (train the Hopfield network). As a result, you obtain the map Tij. Now, add noise to the original patterns by flipping 10% of randomly chosen bits and retrieve the memory by applying the trained network to these. From a multitude of such trials (like 100 or 1000 total), measure the percentage of correctly retrieving the original pattern. The network is considered to work successfully when it retrieves the memory correctly in more than 90% of the cases. Start with only the digits (0-9) and move on to add alphabets if it can successfully memorize (retrieve) all digits. Determine how many different digits or alphabets your network can memorize.

2.Explore the trade-off between the capacity and the noise level, by plotting a 3D graph where x axis represents the number of digits, y axis the noise (flipping) percentage, and z axis the success rate.

3.Repeat the experiments by representing the digits with 1 and -1 instead of 1 and 0. Does this change the accuracy or the convergence rate (in how many cycles the input pattern converges) of the network?

4.Display some of the spurious attractors (the converged pattern that is not part of the memory).

Memorizing the shapes of digits Pattern "0 Pattern I Pattern "2 Pattern e3 Pattern 4 Pattern 6 Pattern . Pattern-9