In my efforts to learn lots of things in a short space of time, I've been thinking a lot about approaches and strategies to learning efficiently rather than writing pages and pages of notes. This is my latest approach…let's call it 'Just in Time Learning'.
Take courses on generally useful topics, but don't study in excessive detail; just ensure that I have at least a basic conceptual understanding. Summarise courses/resources using learning outcomes or similar so I know what information I have, rather than trying to cram everything into my head at once, as this just won't work.
Identify skills gaps in CV and design projects to cover these, with a smaller number of projects spanning a larger number of skills. Ensure projects are about topics that I genuinely find interesting.
Real emphasis on creating a portfolio of work, rather than just learning “stuff”.
Motivation - being a fairly pragmatic person, I'm much more motivated to learn when I have a concrete application for my learning.
Networking - the shift from courses to projects means that I'll have actually tangible projects to discuss with others.
Realism - data is rarely clean and tidy, with lots of tutorials and examples online. Having my own projects will result in having to engage more deeply with problems and come to my own conclusions.
Difficulty - coming up with feasible interesting projects may be difficult, given my lack of expertise.
Frustration - as there are no existing tutorials, I'll have to search harder for the answer when I get stuck.
Isolation - one of the key benefits of MOOCs and Kaggle competitions is having forums and communities to discuss ideas and learn from others.
All that said, it still feels like a good idea. These projects can be done alongside MOOC course projects and Kaggle competitions, and in some ways JiTL mimics the PhD process, which I just about survived intact. I'll post more thoughts of JiTL once I've had more experience practising it :)