Yes. I think it is the best intermediate machine learning book (which can be used also as introductory). More or less same level as Bishop, just that the authors don't hate the reader.
Don't get it confused for Introduction to Statistical Learning (same authors) which is shit.
Duda, Hart and Stork is just not a very good book to be read nowadays. My supervisor loves it but he is ancient. It is really underwhelming and outdated. Same thing goes for Tom Mitchell.
ESL, Bishop and Murphy seem to be the best ones (though as I said, not a big fan of Bishop). Abu-Mostafa is a good one (though different), while I've heard good words about Barber (haven't read it though). After that is either going on more advanced (theoretical) stuff like Shai Ben-David or Vershynin, going in some subclass of ML algorithms (for example Koller for PGM, Goodfellow et al. for Deep Learning though not a big fan of it, Sutton and Barto for Reinforcement Learning), or just going pure math (Michael I. Jordan's list is pretty good, at the very least read Convex Optimizations and Statistical Inference). Of course, books just compliment papers, papers are much more important.
