Hacker News Reader

Comment by quanto | Hacker News Reader

I am looking for a book and resource for (advanced) data science. I understand there are numerous ML books and articles discussing algorithms at all levels; there are also numerous introductory books on data science, e.g. how to call functions in sklearn. What I am looking for is a book on data exploration, data modeling, and prediction error analysis; with some level of rigor in both theory and implementation.

Any recommendation for such a book or resource?

DatabookV2 discussed here looks great for natural science + ML. It reminds me of this widely-cited paper

A high-bias, low-variance introduction to Machine Learning for physicists https://arxiv.org/abs/1803.08823

View on HN

See this MLSS '20 presentation and all recommended bibliography [1], in particular top left box in slide 123.

For rigorous modeling, Gelman et al. books are great [2-3].

[1] https://drive.google.com/file/d/1lPePNMGMEKoaDvxiftc8hcy-rFp...

[2] https://avehtari.github.io/ROS-Examples

[3] http://www.stat.columbia.edu/~gelman/book

Ah, thank you. I did hear about Gelman's books before, but I haven't read them in detail. Will check those out. Do you have references/books for non-Bayesian approach as well?

You could try this one: https://hastie.su.domains/CASI_files/PDF/casi.pdf

However, IMHO many things come more naturally in a Bayesian setting, such as hierarchical models.

Cosma Shalizi's "Advanced Data Analysis from an Elementary Point of View" sounds right up your alley: https://www.stat.cmu.edu/~cshalizi/ADAfaEPoV/

I used it to learn depth on exactly those topics myself. It's just great.

Thank you. Will check it out.

Bishop Pattern Recognition and Machine Learning is an awesome starting point. I think it's way more coherent (although it probably is less comprehensive) than Elements of Statistical Learning.

I do like both of the books; but I found them more focused on mathematical analysis of algorithms rather than analysis of the real world noisy data (and choosing/making which algorithm to use).

Possibly "Learning from Data"[1][2] would be of interest?

[1]: https://www.amazon.com/Learning-Data-Yaser-S-Abu-Mostafa/dp/...

[2]: https://amlbook.com/

https://hastie.su.domains/ElemStatLearn/printings/ESLII_prin...

ESLR then Murphy ML: A PP

ESL is indeed widely recommended for ML but perhaps, I am looking for something more data-exploration focused. In any case, did you find Kevin Murphy's MLPP to be helpful? I read the book in detail and found the book as an introductory mathematical book for discussing ML algorithms, rather than doing any real world data-modeling or prediction error analysis -- it may help with ML, but not sure if it will help me with DS. And the book (at least the edition I used) was full of non-trivial errors.

Reply on news.ycombinator.com