2

How can i prove $f(x_1,\ldots,x_n) = \sum x_i\ln x_i - (\sum x_i )\ln(\sum x_i)$ is convex on R++

Did
  • 279,727
yonigo
  • 129
  • 1
    Please reproduce your question within the body of the question. There are many reasons this is a good practice, but for one example note that LaTeX isn't rendered properly in titles on the android app. – Matt Samuel May 14 '15 at 12:49
  • 1
    Can you add what you have tried? – wythagoras May 14 '15 at 13:03
  • 1
    I am thinking it has something to do with the Jensen inequality, – yonigo May 14 '15 at 13:07
  • 1
    Try taking partial derivatives and prove $\left(\frac{\partial^2 f }{\partial x_i \partial x_j}\right)$ is symmetric and positive semi-definite. – Michael May 14 '15 at 13:37
  • How odd that this was posted twice in one day. http://math.stackexchange.com/questions/1281612/convexity-proof-of-a-function-including-ln-and-sums#comment2602514_1281612 – Michael Grant May 14 '15 at 19:07
  • I'm voting to close this question as off-topic because it is a duplicate. – quid May 14 '15 at 23:27
  • @quid: The other question now has an answer, so I closed this for being a duplicate. – robjohn May 16 '15 at 18:32

0 Answers0