Ask HN: Overwhelmed with learning machine learning, how do I proceed?
Inspired by this thread and the excellent answers (https://news.ycombinator.com/item?id=12882816), I felt courageous enough to open ask myself.
I've completed Andrew Ng course on Coursera but now I feel overwhelmed by the variety of information on machine learning out there. How do I start a new project? How do I choose what problem to pick and what algorithms to apply? Go to https://www.kaggle.com/, pick a project, and apply all algorithms to the problem to see which one is the best. Wow, looks amazing. However I thought of finding my own niche / domain. This seems very structured on the one hand but would prevent me from finding my own small world to apply my new acquired knowledge on. You don't have to stick on Kaggle forever. It is a fantastic resource for learning to apply ML techniques to real life data sets. Spend some time getting good at doing that and then move onto your own domain Here is another site that you might find helpful: https://metacademy.org/ There is a good discussion here: https://news.ycombinator.com/item?id=12713056