I work in Data Science/Deep Learning and I studied Linear Algebra (null spaces, Least Squares, SVD, etc.) at the University, but quite some years have passed and I can't remember all details. Another example of a topic I would like to refresh: when the training set for a machine learning grows in size, not only each training iteration takes longer (this is trivial - more data points to process at each iteration) but also more iterations are needed to get to convergence. I guess this could be related to the condition number of the coefficient matrices of the linear systems we solve during training - bigger matrix size often corresponds to a larger ratio between the maximum and minimum (in absolute value) eigenvalues. I'm an engineer, so I'm looking for a clear and reasonably rigorous reference, but probably not at the level of a math grad student.

Given my goals, is this book a good reference?


If not, can you suggest a better reference?

  • 1
    $\begingroup$ As a start and for application it is a very good book, and you can complete the study with the video lectures. $\endgroup$ – gimusi Jan 22 '18 at 16:37
  • $\begingroup$ @gimusi Thanks, I'll think I'll buy it then. You said it's good for a start: do you think I should complement with something else? I'm not going to do research in the field of Linear Algebra,I mostly need an handy reference while I do research on Deep Learning (there's quite a lot of Linear Algebra involved, and in the near future there will also be a lot of Multilinear Algebra). $\endgroup$ – DeltaIV Jan 22 '18 at 21:39
  • 1
    $\begingroup$ It is a book mainly oriented for an introduction to the topic for engineering applications, it is not a good reference if you want go deeply into the theoretical side. In this case you can refer for example to Serge Lang books. $\endgroup$ – gimusi Jan 22 '18 at 22:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.