4 ms·

This is a great reference, but it’s pretty terse...it would be hard (IMO) to learn something for the first time from EoSL, though it’s great for having all the

This is a great reference, but it’s pretty terse...it would be hard (IMO) to learn something for the first time from EoSL, though it’s great for having all the details and derivations. It’s just a very technical book, and though it is complete it is also quite inaccessible.

thegginthesky 6y ago
That's why they wrote Introduction to Statistical Learning[0] and also a video series for the same book[1]. Both books and the video classes are a must for anyone working with Machine Learning and/or Statistics. [0] http://faculty.marshall.usc.edu/gareth-james/ISL/ http://faculty.marshall.usc.edu/gareth-james/ISL/ [1] https://www.youtube.com/watch?v=5N9V07EIfIg&list=PLOg0ngHtcqbPTlZzRHA2ocQZqB1D_qZ5V https://www.youtube.com/watch?v=5N9V07EIfIg&list=PLOg0ngHtcq...
- michaericalribo 6y ago
  Definitely. And hell, EoSL is a _wide_ survey, and some of the content is only adjacent to ML. Eg, MCMC for Bayesian models: there are good applications in ML, and a short bit in EoSL about the Gibbs sampler, but there’s a massive parallel literature in the statistical inference world.
- jimsparkman 6y ago
  [0] returned a 404, but this appears to be substitute: https://statlearning.com/ https://statlearning.com/
  - thegginthesky 6y ago
    Hmm weird, it works for me. Thanks for the alternative link
- LaserPineapple 6y ago
  And a second edition of ISLR is "coming soon" [1]. [1] https://twitter.com/daniela_witten/status/1261693624439279616?s=19 https://twitter.com/daniela_witten/status/126169362443927961...
kgwgk 6y ago
For a more digestible alternative, see https://news.ycombinator.com/item?id=25592296 https://news.ycombinator.com/item?id=25592296