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