Create machine learning solutions to data science problems by identifying and applying appropriate algorithms and implementations.
We'll be working with the CelebA dataset, a widely used dataset of celebrity faces. The task is to predict the hair colour of the celebrity, which will be one of black, brown, blond or gray. Images are relatively large data items, so we've produced a cut-down version of the dataset.
For testing, there'll be a public test set and a private test set. The public test set is the standard one used for CelebA. The private test set will stay private.
This Jupyter notebook contains all the code you use to train your model(s) and make predictions on the test set, for both the conventional machine learning approach and the deep learning approach. It also contain text blocks with comments explaining what each segment of code does.