Linear Algebra and Optimization for Machine Learning: A Textbook

Absolutely one of my top 3 technical book purchases for mathematics in machine learning and data science, out of my approximately 20 books total on these topics. Aggarwal has constructed a text that is both mathematically clear and with a clearly-written progression to more advanced topics in Linear Algebra. I especially purchased this book for more mathematical rigor *and* applications focus in Eigenvectors, Eigendecomposition, Principal Component Analysis, and Singular Value Decomposition; and Aggarwal has built an admirable construction covering these topics. Reference my photos of the start of the “Applications of Singular Value Decomposition” sections. (This section occurs about the 2/3 of the way through the book; so another 1/3 of the book’s content follows.) Highly recommended.

A very fine quality book, looks and smells great 😁 It is affordable as well. Subject review: Just started with the subject. Whatever little pickings I did, felt impressive. I plan to come back and do a thorough review after burning a few weeks/months of midnight oil!

This book is very hard for me to understand. I need to read it very slowly and keep reading the same sections multiple times before it makes sense. There are very few worked out examples in the book. There are no answers to any practice exercises in the book. There are very few diagrams, and I feel like it would greatly help in explaining some concepts. Often times the author states a conclusion or consequence of some lemma, but I have no idea how that conclusion follows from the equation or lemma. Most of the practice problems and exercises are very abstract in nature, and call for proving something. For example, if a matrix has quality A, then prove that it also has quality B. These problems don't really help me to understand the material. I was expecting more practical problems like: invert this matrix, decompose this matrix, find the orthonormal basis of this matrix, find the determinant of this matrix, etc. Those sort of problems are rare, and often times I feel like the exercises don't align with the chapter's content. Problems with actual numbers could really help me to better understand the material. On the plus side, it seems to cover much more material than a standard textbook does. It could easily have been split into two or even three distinct books covering the same content. Unfortunately that means information is crammed into this and hence tough to comprehend.

I found the book very useful. It provides a very good coverage of all the background in linear algebra and optimization needed to understand machine learning papers and tools, without having to read separate books on these two topics. This includes specialized topics needed in machine learning, which are not found in introductory books on these topics. There are a lot of good exercises in each chapter, some of which are pretty challenging, and these are very useful in getting a good grasp of the material.

Like the proofs

I've read more than a few books on both linear algebra and machine learning and this one is by far the best. Concise explanations that make the concepts very clear. Almost like Gilbert Strang in his clear explanations of the concepts

I’m currently studying machine learning and the math in this book really fits my needs.

Doesn’t provide solutions to check your learning. I am on here trying to independently learn, not buying books to go to overpriced college classes.

Superb book. Condition in which the book came was not good.

Llego de manera excelente y es un gran libro ya que muchos problemas del machine learning surgen de problemas de optimización.

On pourrait croire que c'est livre pour débutant, on explique qu'est-ce qu'une matrice, un vecteur, un scalaire, les types de matrices (rectangle, triangulaire etc) mais les exercices sont d'une difficulté ! Évidement, il n'y a pas de correction avec le livre. Le livre est rempli de lemme et preuve, un style trop académique à mon goût. C'est effectivement un livre qui se concentre sur le machine learning Je le lis quand même, ce n'est pas non plus un mauvais livre, mais un livre de révision de l’algèbre linéaire et de l’optimisation que d’apprentissage

Livro muito bom. Atendeu plenamente às minhas expectativas do momento. Além disso, poderá ser útil, futuramente, no desafio de preparar aulas de Álgebra Linear/Otimização, no contexto de um programa relativamente extenso, de maneira a atender aos objetivos do curso.

Livro tem conteúdo muito bom. Mas veio danificado e eu tentei trocar mas perderam duas vezes o livro e só quando o terceiro chegou veio novamente danificado. Provavelmente veio agora o mesmo livro que devolvi.