Cloud Atlas is a deep learning algorithm designed to predict the appearance of shallow, climate-impacting cloud patterns among satellite images via convolutional neural networks. This is a current Kaggle competition and represents my first attempt at multi-label image classification.
The Kaggle competition link is: https://www.kaggle.com/c/understanding_cloud_organization/overview
The presentation link is: https://docs.google.com/presentation/d/18unKVvFk4EZn8iV5Us75uFTdDAzwxt8nMnNyAjak-_g/edit?usp=sharing
The code is available in this repository.