Skip to content

zhuhs-q/basic-optimal-method

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

32 Commits
 
 
 
 
 
 
 
 

Repository files navigation

basic-optimal-method

  • the project is intended to collect basic convex optimal methods and their extensions.
  • they are raw and of course not for commercial use. most of them haven't been packed yet.
  • for me, the project is for reviewing those algorithms, and i hope it may help you in some ways.
  • every class shall be updated within a week

descent-method

  • In fact, all the descent algorithms are examples of it, the difference is the norm they use
  • for convenience, all the descent algorithms equiped with backtracking line searching
  • examine it with "descent_method_main.jl"

gradient descent method

  • example of steepest descent method with l2 norm
  • use backtracking for line searching
  • not packed yet
  • example : x' * x

coordinate descent method

  • example of steepest descent method with l1 norm
  • use backtracking for line searching
  • not packed yet
  • example : x' * x

newton method

  • example of steepest descent method with hessian matrix
  • not packed yet
  • example : x' * x

Linear programming

Linear programming is never easy programming, clever trick and geomerical intuition lead us to a higher level

simplex method

  • in my opinion, the simplex consists of two stage: find the initial point and find the optimal point. and in the program, these two stages are based on geometrical intuition and easy to understand.

simplex method with cutting for Interger Programming

  • gomory's cutting strategy