The MNIST neural network was used to check if the neural network was learning before implementing it into the convolutional neural network later used in the project.
The neural network was able to be trained on a dataset and got a accuracy of 95% which is quite decent.