From the course: Neural Networks and Convolutional Neural Networks Essential Training
Join today to access over 25,300 courses taught by industry experts.
Testing the neural network model
From the course: Neural Networks and Convolutional Neural Networks Essential Training
Testing the neural network model
In the last video, we only checked how well our neural network performed on a few sample images. But to really understand its accuracy, we should run it on the entire test dataset. Now before we do that, it's super important to remember the difference between the training set and the test set. The training set is the one we fed into the model in batches of 64, so that's 30 times in total, to help it learn and to adjust its weights. Our model has only seen a couple of images from the test set, but it hasn't been trained on them. So the images in the test set will not be the same as in the training data set. Otherwise, the model could just regurgitate whatever patterns it's memorized. So to be clear, both the training set and the test set have images from the same 10 clothing classes, but they contain different images. So let's head over to Google Colab. So the first thing we want to do is to use our test loader. That'll allow us to load batches of 64 images for testing. We defined our…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.