The books "The Elements of Statistical Learning" by Trevor Hastie, and "Advances in Financial Machine Learning" by Lopez De Prado are highly recommended books for ML. They both deal with machine learning algorithms, and the statistics involved in ML algorithms. De Prado's book deals with ML for finance while Trevor Hastie's book seems to be generic. However, if ML for finance is concerned, which book is preferred and what are other differences between these books if any?
$\begingroup$
$\endgroup$
4
asked Apr 24, 2022 at 18:30
-
2$\begingroup$ Hi: you're probably not getting any answers because you answered the question yourself. Haste et al is more generic than De Prado which is specific to ML in finance. There's really not much else to say because it depends on what you want to learn. Personally, I would get both if possible. $\endgroup$– mark leedsCommented Apr 25, 2022 at 17:01
-
$\begingroup$ @markleeds Thank you for the response! I appreciate you leaving a helpful comment in response to my question. I agree with you in that I should get them both. I will do that. Thanks once again! $\endgroup$– TryingHardToBecomeAGoodPrSlvrCommented Apr 25, 2022 at 17:49
-
1$\begingroup$ Assuming you get both, I would go through them in the order of Hastie et al first and De Prado second. Better to get the big picture first IMHO. $\endgroup$– mark leedsCommented Apr 26, 2022 at 4:26
-
$\begingroup$ Right. ESL talks about the nuts and bolts of statistics involved in various ML algorithms and AFML talks more about how to apply various ML techniques effectively in trading. So it does make sense to study them in that order. Thanks once again for the helpful suggestions and comments! $\endgroup$– TryingHardToBecomeAGoodPrSlvrCommented Apr 26, 2022 at 7:50
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
1
As mentioned in your question, that "if ML for finance" is concerned, then I think De Prado's book should be preferred, since his book puts more emphasis on how to apply data science techniques to actual problems in finance. However, even with "Machine Learning" in the title, the book actually deals more with data analytics rather than concrete machine learning algorithms, so you might want to take that into consideration. Nonetheless, the book still outlines the basics of a prediction task: it tells you how to correctly define a prediction label, shows you some features to build a model, and how to use your prediction signals to trading. So at least you get a general framework of the process.
ESL covers a broad variety of topics and some of them are not practical in the financial domain. I would suggest you use it as a reference book when stuck with certain specific algorithms. If you are just starting to learn machine learning, then there's a simple version of ESL called An Introduction to ESL which you could quickly finish and move on to AFML. If you are already familiar with ML, just directly read Prado's book.
-
$\begingroup$ Thank you for the nice answer. I had heard about ISL and the first author of ISL mentions the same thing as you do. The content in Hastie's ESL is perhaps a bit of an overkill for a practitioner. Your answer confirms my understanding about the content of these books, and about the approach I was planning on taking. Thanks once again! $\endgroup$ Commented Jun 6, 2022 at 11:52