Its very easy to get overwhelmed by the volume of materials/resources available on internet these days to study Neural networks and Machine learning.
When I started, I could not decide what to learn from where. There were multiple excellent courses/blogs/tutorials which I felt to be 'must do'. This information overload made me end up doing nothing at worst or wasting a lot of time to decide which one to pick at best.
A few times I stumbled upon some suggestions of how to structure my learning and handle this information overload. This blog contains list to such resources which proved invaluable to me to structure my learning and move fast.
1. This describes how to structure your project into tasks and sub tasks and allocate appropriate time resources to them:
2. Majority of us don't get the opportunity to attend courses in an Ivy league university in-person. I have learned a lot from the lectures of courses hosted on MIT OCW and Stanford online. Andrej Karpathy's advice on how to do well in your courses teaches a lot things about human memory and work ethics you should follow while doing these courses. Most importantly, experiment what works best for you instead of just following all of the points.
3. Tips from writing a good paper to finding a compatible advisor to structuring your approach for navigating through research ideas and other important guidelines, this article from Andrej Karpathy covers a lot things you will need to survive in a research community. You need not be thinking to go for a PhD in order to use this valuable discussion.