Here are some useful resources.

Google Colab

Google colab is free cloud computing for python that supports GPU computation and includes tensorflow by default.

Google Machine Learning Crash Course

This is free introductory course to machine learning that should take around 15 hours to complete.

The Hundred-Page Machine Learning Book

This book covers the basics of machine learning techniques in a compact format (just over a hundred pages). The book was written by Andriy Burkov and went through an open online review process where the draft of each chapter was available for free for anyone to comment on. The PDFs are still available, and almost match the book word for word, so are worth a look if you want to brush up on some of the fundamentals.

An Interactive Node-Link Visualization of Convolutional Neural Networks

A great tool for understanding how data propogates through a neural network. You can draw your own inputs and see which weights effect the output.

Intro2ML

A collection of juypter notebooks that serve as examples for beginners of using neural networks to tackle different problems. The neural networks are written in Keras.

NOTE: This is very much still a work in progress and in the early stages of development. Any feedback/contributions are much appreciated.

Machine Learning Mastery

Machine learning mastery has quick-start guides and blog posts on a variety of topics in machine learning that aim to get straight to the point and help you understand core ideas.

Tensorflow playground

A useful tool to help understand the basic concept on a neural work. The user can tinker with a neural network and see how the output changes.

Tutorials