How to solve min-max problem in CVX?

#1

A similar question has been asked here
http://ask.cvxr.com/t/is-it-ok-to-solve-min-max-problem-by-cvx/942?u=user01

I am wondering, how could I define the auxiliary variable to solve this min-max problem in CVX? Is there any particular rule to be followed in CVX?

(Mark L. Stone) #2

The chance of getting a useful answer will be increased if you completely describe a sample problem of interest.

You may not even need an auxiliary variable.
max of convex expression arguments is a convex expression which can be minimized. For instance, this made-up (useless) problem is valid CVX code:

cvx_begin
variable x(4)
minimize(max(x.^2))
cvx_end
#3

Thank you. This helps. I will post the MWE once I complete the code.