Andrew Ng lecture, Stanford Graduate School of Business
High level, but totally worth watching.
https://www.youtube.com/watch?v=21EiKfQYZXc
I went looking for lectures from different sources and these are a great "from first principles" introduction. The newer cs231n lectures may have more relevant details, but I like Andrej Karpathy's style.
Lectures: https://www.youtube.com/watch?v=NfnWJUyUJYU&list=PLkt2uSq6rBVctENoVBg1TpCC7OQi31AlC
Website: http://cs231n.stanford.edu/
Notes/Code: http://cs231n.github.io/
"The book" on deep learning, but assumes a very high competency in statistics and is not super approachable. That said, I would still recommend it if you enjoy reading "the book" and because it covers fundamentals I haven't seen elsewhere.
https://www.deeplearningbook.org/
I'm not going to recommend specific papers, but I found it very useful to read through papers even as I just started learning. And if you're looking for research, checkout arxiv-sanity!
Open source web site developed by Andrej Karpathy to help filter ML research with the goal
of finding more relevant research faster. I like the twitter feature, top hype :)
http://arxiv-sanity.com/
To start, there are really only two you should choose from: Tensor Flow and PyTorch. Having looked at both of them, I would strongly recommend PyTorch, both because of how much easier it is to get started and because it's favored by the research community.
That said, if you're going into production, Tensor Flow is the better choice, as it scales from multi-cpu, to GPU, to cloud, down to phones.
If you do go with PyTorch, check out the examples:
https://github.com/pytorch/examples
The micro-lectures are nice and the "colab" exercises are a really easy way to get started on the code side, Tensor Flow exclusive though.
https://developers.google.com/machine-learning/crash-course/
Can be a little silly, but he has a great stream of bite-sized ML content that is perfect for idle moments.
https://www.youtube.com/channel/UCWN3xxRkmTPmbKwht9FuE5A
Not ML-specific and not super educational, but skews heavily toward machine learning content and offers great little sumaries of current research.
https://www.youtube.com/channel/UCbfYPyITQ-7l4upoX8nvctg
I haven't checked these out, but many people say they are great learning resources.
Also haven't read these, but I hear they are much more approachable than the "Deep Learning" book by Goodfellow.
https://www.amazon.com/gp/product/B079XSQNRX/ref=as_li_qf_asin_il_tl?ie=UTF8&tag=andrewglass0a-20&creative=9325&linkCode=as2&creativeASIN=B079XSQNRX&linkId=f62bf347940f2e95e3907c7d15cf16b5
Designed to be as accessible as possible (e.g., Deep Learning For Dummies)
https://www.youtube.com/watch?v=r0Ogt-q956I
Light on the theory, heavy on the application, written by François Chollet, creator of Keras.
https://www.amazon.co.uk/Deep-Learning-Python-Francois-Chollet/dp/1617294438