I have never taken part in an official or structured data analysis or machine learning course (other than recent online offerings) and have learned most of what I know from reading and trying things out. I know I am far away from being able to get a job.
My question isn’t what is better (like this question) but rather, can I reach a level where I can apply for a job and actually have a chance with self-study alone? Also, is it possible to do this within a reasonable time frame (maybe 10 years? I’m 31 now so…)?
Or will I have to find a way to attend some sort of college/uni?
It’s all about being able to show a potential employer that you have the skills they are looking for. A degree from a college is one piece of information that an employer can use for that, but not the only thing (nor does it necessarily translate into real world skills).
For me as a hiring manager even more important than that is experience and hands on examples. If you want to work in data analysis or machine learning my advice to you would be to do as much data analysis and machine learning work as you can. Start a blog, open a Github account, compete in competitions like on Kaggle. Depending on where you live, find a meetup, hackathon, etc.
Not only will you learn a lot from those experiences, you’ll also meet a lot of people in the field and generate some examples of work that you can show an employer.