Stanford EE364A Convex Optimization I Stephen Boyd I 2023 I Lecture 6

[@MinaKh-s5h]

I didn't get affine and differece between affine and linear?

[@JiataiTong]

piecewise linear does not guarantee convex / affine. Actually it is affine when you have a piecewise affine function composed by the max of a bunch of affine function. However, you may see lots of other linear spline functions like ReLU network that are not convex/affine. And you need mixed integer linear programming to solve that kind of tasks (introduce integer variables to denote which piece of function you are at).

[@ML_n00b]

Didn’t really understand the quasi convex sub level set arguments from before

[@floribertjackalope2606]

You find a feasible point of a convex function phi(x) s.t. f(x)

[@Maxwellpaulwall]

"warning warning may not be LP ahead" 🤣

[@ddxoxbb9106]

🤣🤣🤣