$$\min_{x \in \mathbb{R}^n} \max_{1 \leq i \leq k} \frac 1 2 x^T Q_i x + b^T_i x + c_i$$ $$Ax = d$$

$Q_i$ are positive definite matrices. Find dual problem.

I found out that this problem called quasiconvex and can be solved by moving "max" part into constraints section. So I did this:

$$\min \gamma$$ $$\frac 1 2 x^T Q_i x + b^T_i x + c_i \leq \gamma$$ $$Ax = d$$

But there are two things I am confused about:

  • The constraint is not linear
  • I am not sure about how to minimise $\gamma$
  • $\begingroup$ The original problem is convex, not just quasiconvex. You minimize $\gamma$ with a QCQP solver. $\endgroup$ – LinAlg Nov 1 '19 at 12:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.