Deep Learning Studies

This repository will show various implementations of DL techniques using TensorFlow 2.0.

1. Convolutional Neural Network (CNN)

The first project will be the creation and training of CNN to classify images of the CIFAR 10 dataset.

• A deeper model was used to raise val_accuracy to ~85% after 25 epochs.
• Some data augmentation such as small rotations and zoom-ins raised the accuracy even further.