Simple Classifier Performance

Trick to analyze!

A simple way to see how well our neural network correctly classifies images is to work through all the 10,000 images in the MNIST test dataset and keep a count of how many it got right. We do this by comparing the output of the network with the label that came with the image.

The following code sets a score count to 0 and then works through the test data, incrementing the score every time the label matches the network’s output.