| I can list a bunch. But the answer depends on a lot of things. 1. Which language? 2. Do you have any background in programming and if yes how much and in which language? Beginners can range from "I don't know what a loop is" to "I've been a front end developer for a few years and want to try something new". These people will obviously need to approach things very differently. 3. What's your math background? Do you know what a derivative is and how matrix multiplication works? Do you want to go in depth or do you need just a general understanding of the algorithms? Some people will need to start at precalculus if they really want a solid foundation. The same people also most likely won't have the patience or interest to stick with the theory for long enough. 4. How will you use what you're learning? Is there a specific goal related to this and does it have a time horizon? ... Anyway, I'm rambling. Data science is just programming and statistics. Your general learning lanes are 1. Theory - calculus, linear algebra, statistics, ML algorithms. 2. Programming. 2a. Good software development principles - writing maintainable code, version control, testing, design patterns etc. 2b. Tooling - learning the language-specific ecosystem of libraries. This is what most beginner resources (including the OP) focus on and is also the one that constantly expires and has the least transferability to other skillsets and fields. Obviously, it's still necessary - using the right tools and knowing them well goes a long way. You mentioned ISLR - it's a good beginner-ish theory book that helps you understand how the algorithms work without going too deep into the math. |