How to implement a convex optimization package?

I fully understand that Convex Optimization packages, such as Linear Algebra packages, should be used by you, not implemented. However, for purely educational purposes - is there a good resource - a link / book on how to implement the convex optimization package? (for example, for quadratic programs with quadratic constraints?)

Thank!

+3
source share
3 answers

Any good convex optimization tutorial has the things you are looking for. One such free, but great resource here is: CO Book . Note that, as you say correctly, the implementation of the algorithms mentioned in this book will definitely require linear algebra libraries that you may or may not want to implement.

+2
source

The Optima has an article, newsletter Society mathematical optimization called "rapid development Minlp Solver open source with the COIN-OR". It describes the construction of a nonlinear solver using some coin-or packages . Most of the coin or material is written in C ++.

python , numpy. scipy , .

, Boyd GPL python cvxopt, , .

+1

, , , . , , .

, (http://cs229.stanford.edu/proj2017/) - cvx4ml, , SkLearn, 24- , , :

, . :

  • /
  • /
  • 20 ( , )
  • , , .. , .
  • , , LU, Cholesky.
  • . , . 6.a - . 6.b - 6.c - .

  • "5"

CVXPY,

  1. , CVXPY, .

p.s. , :

  • , .

  • youtube EE263 S.Boyd, EE364A S.Boyd, EE364B S.Boyd.

0
source

Source: https://habr.com/ru/post/1753621/


All Articles