l**********r
发帖数: 1 | 1 Thanks for your help. Looking forward to your expert opinion.
I have an optimization problem as follows
(P_0) min f(x)-g_{i^*}(x)
s.t. i^* = arg min_{i=0,\ldots,k} g_i(x)+e_i
where e_i is a constant only dependent on i. f and g_i (x) are not
necessarily convex.
To avoid discrete functions (or integer variables), I use this new
formulation
(P_1) min f(x) - z
s.t. z \leqslant g_i(x) + e_i, \forall i=0, \ldots,k
Can I find the (global) optimal solution x^* of (P_0) by solving P_1? (for |
|
