Prob/Stats Any good math-theory-focused books on neural networks and data science?

AI Thread Summary
The discussion centers on finding books that provide a mathematically rigorous approach to data science, particularly in the context of neural networks and optimization theorems. Recommendations include "Algorithms for Optimization" by Kochenderfer, which features practical examples in Julia, and "The Hundred-Page Machine Learning Book" by Burkov, available as a try-and-buy online option. Additionally, "Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow" by Geron is noted for its mathematical discussions, though it may not meet the desired level of rigor. The conversation acknowledges that many rigorous insights are currently found in academic papers rather than textbooks. Several relevant research papers on function approximation, gradient descent effectiveness, and error estimation are also shared, indicating ongoing advancements in the field.
s00mb
Messages
33
Reaction score
10
Hi. I'm looking for books on data science, preferably leaning towards neural networks, that focus on mathematical rigor. For example, theorems on optimization, minimum number of layers to accomplish a task efficiently, etc. Most books I've seen seem to hand wave this stuff. Anyone know any juicy books on the topic?
 
Physics news on Phys.org
There's a rather recent book by Kochenderfer called Algorithms for Optimization with many examples written in Julia, a hot programming language from MIT that folks are using for numerical work in diverse fields including ML and Data Science.

https://www.amazon.com/dp/0262039427/?tag=pfamazon01-20

There's also the 100 page ML book by Burkov:

https://www.amazon.com/dp/199957950X/?tag=pfamazon01-20

which is available online as a try and buy book.

Lastly, Geron's book Hands-on ML with Scikit-Learn, Keras and Tensorflow:

https://www.amazon.com/dp/1492032646/?tag=pfamazon01-20

All are good books that discuss the math behind the ML although not at the rigor you're looking for.
 
I looked through "Data mining" by Witten, Frank, Hall and, Pal which covers most of these, but it isn't rigorous like the good old math analysis books. No wonder for such a rapidly developing (practical) area.
 
TLDR: is Blennow "Mathematical Methods for Physics and Engineering" a good follow-up to Altland "Mathematics for physicists"? Hello everybody, returning to physics after 30-something years, I felt the need to brush up my maths first. It took me 6 months and I'm currently more than half way through the Altland "Mathematics for physicists" book, covering the math for undergraduate studies at the right level of sophystication, most of which I howewer already knew (being an aerospace engineer)...
I've gone through the Standard turbulence textbooks such as Pope's Turbulent Flows and Wilcox' Turbulent modelling for CFD which mostly Covers RANS and the closure models. I want to jump more into DNS but most of the work i've been able to come across is too "practical" and not much explanation of the theory behind it. I wonder if there is a book that takes a theoretical approach to Turbulence starting from the full Navier Stokes Equations and developing from there, instead of jumping from...
Back
Top