Introduction¶

In this assignment, we are going to build, train and validate convolutional neural networks (ConvNets). For this purpose, we will use data from the publicly available CIFAR10 dataset. CIFAR10 is a dataset commonly used in the community of computer vision and machine learning to benchmark new algorithms and network architectures.

CIFAR10 is a dataset that contains (small) RGB images of 32x32 px of ten different classes:

1. airplane
2. automobile 
3. bird
4. cat
5. deer 
6. dog
7. frog
8. horse
9. ship
10. truck

More details can be found at this link: https://www.cs.toronto.edu/~kriz/cifar.html