The MNIST database of handwritten digits contains 60,000 images of digits that can be used to train a model and 10,000 images that can be used to evaluate its performance. It is a great dataset for evaluating algorithms on the problem of handwriting digit recognition.
Great results can be achieved on the MNIST problem using convolutional neural networks (CNNs). In the notebook, we build a simple CNN comprised of one convolutional layer, one max pooling layer and one dense layer to make predictions. The CNN is first trained and its classification accuracy is calculated on the test set.