I have an optimization problem with the following objective function in $(x,y)$ $$ A\log \left(\sum_{i=1}^n x_i\right)+\log\left(1-\frac{f}{n}\left(\sum_{i=1}^n\frac{x_i}{y_i}\right)\right) $$ where $x_i\geq 0$, $y_i\geq x_i$, and where $0<A<1$ and $0<f<1$ are constant. I would like to use Mathematica to verify its convexity.

I can compute the Hessian. Let's focus on the case with $n=2$:

F[x1_, x2_, y1_, y2_] = Log[x1 + x2] + Log[1 - 0.1*(x1/y1 + x2/y2)]
A := D[F[x1, x2, y1, y2], {{x1, x2, y1, y2}, 2}]

I can then compute the eigenvalues of A but they are a gigantic mess. Is there anyway I can test whether the eigenvalues are negative for a set of $(x,y)$? Any help would be greatly appreciated!


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.