// The Top 3 Books to Learn Math for Data Science Right Now · Learning With Data

The Top 3 Books to Learn Math for Data Science Right Now

Dec 7, 2019 14:09 · 501 words · 3 minute read books math data science

Photo by [Antoine Dautry](https://unsplash.com/@antoine1003?utm_source=medium&utm_medium=referral) on [Unsplash](https://unsplash.com?utm_source=medium&utm_medium=referral) Photo by Antoine Dautry on Unsplash

Awhile back I wrote an article on the top books to get started with data science:

The Top 3 Books to Get Started with Data Science Right Now
_And build a great foundation of knowledge_towardsdatascience.com
[](https://towardsdatascience.com/the-top-3-books-to-get-started-with-data-science-right-now-3f9945d78011)

One of the readers left a comment asking for the best books to learn math necessary for data science. I thought that was a wonderful idea! If you are eager to strengthen your mathematical foundation and really understand the inner workings of machine learning algorithms, this will give you a great start!

Introduction to Linear Algebra

Linear algebra is core to understanding most of today’s machine learning algorithms. In my opinion, there is no better introductory text on linear algebra than Gilbert Strang’s Introduction to Linear Algebra. What other linear algebra book has 4 stars and over 100 ratings on Amazon? Plus, you can get his course online for free via MIT’s open courseware. If you still struggle to understand the linear algebra of machine learning, then look no further than this book to build your knowledge. Strang is an excellent teacher and his course covers topics such as least squares, eigenvalues/eigenvectors, and singular value decomposition.

The Matrix Calculus You Need For Deep Learning

From fast.ai’s Jeremey Howard, who strives to make deep learning approachable, comes a great “book” that covers all the matrix calculus necessary for deep learning. The goal of this paper is to, “explain all the matrix calculus you need in order to understand the training of deep neural networks.” I think it does a great job and I have yet to find anything as approachable and focused on the calculus necessary for deep learning. If you need a refresher on the basics of calculus, check out the introductory book from Gilbert Strang on the subject. It’s free!

Doing Bayesian Data Analysis

Once you’ve got linear algebra and calculus down, its time to move onto statistics. There are many ways in which you might learn the foundations of stats, but my favorite way is to focus on Bayesian statistics. Bayesian methods will force you to really understand probability and sampling. My favorite book for this area is Doing Bayesian Data Analysis. This book does an amazing job of starting with the basics and building to advanced topics. What is even better are the included examples with data and code! The examples are in R. If you want some similar examples but in Python, check out Bayesian Methods for Hackers.

Bonus

Now that you understand linear algebra, calculus, and statistics check out the Deep Learning book. Not only will it provide a refresher on the mathematics, but it will also show you how all the math connects to make deep learning algorithms work.

Hopefully, you find these books as helpful as I have. If you take the time to really understand the concepts they cover, you will be well on your way to truly understanding how machine learning algorithms work.

Join my email list to stay in touch.

LinkedIn Share tweet Share