I have finished the first two lectures. What past Kaggle competitions would you suggest that I attempt that are like the Dog vs Cat, DogBreeds competitions.
I feel trying to work with the different datasets will really help me understand what I am doing rather than just going through the notebook.
Thanks
Ali
The Fruits-360 dataset was fairly similar to the Dog Breeds dataset. I liked it because not a ton of people have done it and there aren’t so many kernels – which in my opinion is good because it forces you to not rely on other people’s models/solutions (but there is some help when you need it). Also, you can get nearly perfect accuracy (~99%) with some good fine-tuning 
Plant Seedlings is a nice little follow on.