| If you know almost nothing about the field, then introduction to statistical learning is a good choice. http://www-bcf.usc.edu/~gareth/ISL/ISLR%20First%20Printing.p... It assumes some understanding of calculus, but doesn't require matrix algebra. The original (and amazing) book that lots of people used is Elements of Statistical Learning. https://web.stanford.edu/~hastie/local.ftp/Springer/OLD/ESLI... Chapters 1-7 are worth their weight in gold. This is one of the cases where the physical books are much better, as you'll need to flick back and forth to see the figures (which are one of the best parts). The forgoing assumes that you already know some statistics/data analysis (the latter probably being more important). If you haven't done this before, then I suggest that you acquire some data you care about, install R (a good book is the Art of R Programming by Matloff), and start trying to make inferences. And draw graphs. Many, many, many graphs. If you keep at this, finding papers/books and reading theory, and implementing it in your spare time, then you can probably get a good data science job in 1-2 years. You'll probably need to devote much of your free time to it though. I'm assuming that you can already code, given the context :) |