My advice: Do not label yourself as a data scientist or machine learning expert. Go for the domain, i.e. become comfortable with the actual data and the method

My advice: Do not label yourself as a data scientist or machine learning expert. Go for the domain, i.e. become comfortable with the actual data and the methods used there: - predict land use in aerial imagery - become comfortable with photogrammetry, geography, etc. - predict biological tissue(s) - become comfortable with specific branches of biology or medicine - predict $something_relevant I actually stole this advice from the epilogue of some text about programming, and it really stuck with me. Otherwise your expertise is just too generic and you compete with a big pool of people who call themselves machine learning experts, because they can write a for loop in Bash.