Deep learning is a family of computational methods that solve problems that sound intuitive when stated, like recognising a face in an image, but where it's hard to explicitly write down all the steps a computer needs to follow to solve them. These methods work by allowing computers to learn from experience, whereby they build an understanding of the world in terms of a hierarchy of increasingly sophisticated concepts. Deep learning methods have dramatically improved the state-of-the-art in many application domains: from voice-based virtual assistants, to dermatologist-level classification of skin cancer, to self-driving automobiles.
This repository offers the following:
- A systematic pathway to learn deep learning, incorporating the best resources on the web.
- A clear guide through the state of the art of deep learning applications across different domains.
It is a work in progress. Check back in frequently for updates.
Harish Narayanan, 2017