Cool Machine Learning Books(matpalm.com)
matpalm.com
Cool Machine Learning Books
http://matpalm.com/blog/cool_machine_learning_books/
16 comments
Agreed that the reviews are not in-depth. But sometimes inclusion on the list is enough.
I saw enough books on the list that I recognized and benefited from to convince me that the author has some idea of what they're talking about. That's enough to convince me that some of the other books on the list might be worth checking out.
Once you have a book title in mind, there are lots of resources you can use to find out whether it's worth reading -- Amazon comments, book reviews, the publisher's website, the table of contents, published excerpts, etc etc etc.
The most important part of a book search on the internet today is title discovery: picking one book to look at out of the gazillion books available. The author seems to have taken a pretty good stab at that.
I saw enough books on the list that I recognized and benefited from to convince me that the author has some idea of what they're talking about. That's enough to convince me that some of the other books on the list might be worth checking out.
Once you have a book title in mind, there are lots of resources you can use to find out whether it's worth reading -- Amazon comments, book reviews, the publisher's website, the table of contents, published excerpts, etc etc etc.
The most important part of a book search on the internet today is title discovery: picking one book to look at out of the gazillion books available. The author seems to have taken a pretty good stab at that.
But it's so hard for anyone who's read Gilbert Strang's books to refrain from superlatives when describing them!
Strang is magnificent
Lectures: https://ocw.mit.edu/courses/mathematics/18-06-linear-algebra...
Lectures: https://ocw.mit.edu/courses/mathematics/18-06-linear-algebra...
I'd vouch for MacKay's book. It is a real gem and it gives you many big picture insights, as well as specific computations. Moreover, you can download the .tex sources and read the comments (including variants of some difficult paragraphs), there's a lot of snark in them!
100% Agreed! I wonder if anyone has actually made the .tex compile. Last time I tried, I tried to get rid of/define empty versions of missing macros, images, etc. but it was a bit beyond my texing ability.
I've always wanted to be able to "annotate" a textbook I was reading ::in-place::, and write my own "local" version of the text.
I've always wanted to be able to "annotate" a textbook I was reading ::in-place::, and write my own "local" version of the text.
I didn't know about this, great way to get better insight
> . i still feel like evolutionary approaches are due for a big big comeback any time soon....
I was in college during the early late 90s, early 2000s, that was a winter time for ML.Besides the perennial AI, the popular term was "soft computing" to include not only neural networks, but also stuff like fuzzy logic, evolutionary algorithms and simulated annealing. I found it all the approaches so fascinating. For better or for worse the DL explosion totally eclipsed the rest of approaches. Does anybody know of recent relevant work on those fields?
I was in college during the early late 90s, early 2000s, that was a winter time for ML.Besides the perennial AI, the popular term was "soft computing" to include not only neural networks, but also stuff like fuzzy logic, evolutionary algorithms and simulated annealing. I found it all the approaches so fascinating. For better or for worse the DL explosion totally eclipsed the rest of approaches. Does anybody know of recent relevant work on those fields?
There’s some cool stuff from OpenAI and Uber AI Labs scaling up neuroevolution (disclosure: these were my former colleagues at Uber). Check out the work from Ken Stanley and Jeff Clune, plus Tim Salimans. They solved most of the Atari suite comparably to the DQN from Deepmind (these are RL tasks).
Thanks a lot, I will definitely check them out
Here's another book I recommend: Course in Machine Learning by Hal Daumé III: http://ciml.info/ , pdf at http://ciml.info/dl/v0_99/ciml-v0_99-all.pdf
+1 for Speech and Language Processing.
Not only helped me with NLP but really solidified backprop.
I'd love a machine learning books and not machine-learning books.
The site has been suspended
I also recommend Python Data Science - OReilly.
Nice list
That said, most of the descriptions on this page are not detailed enough to tell me anything useful about them. There is little specific about, say, author perspectives, level of formalism vs appeals to intuition, interesting proofs or explanations of particular theorems/topics, helpful or not indexes, etc.
Saying "X book is great" or "I liked book X" and not giving a reason does nothing for me. As written, these reviews detract from the list because they draw the eye to no benefit. As a rule of thumb, the words "good", "great", and "epic" should be avoided at all costs in this kind of writing
I apologize for being harsh -- I'm giving this feedback because I didn't notice any affiliate links which suggests to me that you do want for this to be a useful resource (not that it wouldn't be true if it had affiliate links, just would suggest other motivations took greater priority)
Thanks for spending the time on it regardless