Let $f,g: \mathbb{R}\rightarrow \mathbb{R}$ be two strictly convex functions, where $f$ is differentiable, $g$ is smooth, and $f\geq g$. Suppose that for some $x_0\in \mathbb{R}$:

$f(x_0)=g(x_0)\hspace{10mm} f'(x_0)=g'(x_0)$,$\hspace{10mm}$ and $g''(x_0)>0$

Does there exists an interval $I$ of $x_0$ such that $f$ is strongly convex on $I$?

Note: There does exist such an interval for $g$.


No, there need not be such an interval. Take $g(x)=x^2$ and $f_0(x)=2x^2$ (this is not yet $f$). Choose a sequence of disjoint intervals converging to $0$, for example $I_n=[2^{-2n},2^{1-2n}]$. Redefine $f_0$ to be linear on each $I_n$; this creates a function $f_1$ which is not strictly convex in any neighborhood of $0$, and still satisfies $f_1\ge g$. But you want $f$ to be strictly convex, so let $f(x)=f_1(x)+x^4$.

It's true that $f_1$, as written above, is not differentiable. To fix this, replace the line segment over $I_n$ with the following three-piece gadget:

three pieces

Namely, the middle third of line segment is brought down a bit (it stays above the graph of $f_0$) and smooth transitions are added on left and right.

  • $\begingroup$ Thanks for great example, but $f_1$ isn't differentiable, is it? I get $f'_1(1/4)=1$ from left and $f'_1(1/4)=3/2$ from right. $\endgroup$ – Ashley Aug 8 '14 at 18:35
  • $\begingroup$ @Pilo Right; see my edit. $\endgroup$ – user147263 Aug 8 '14 at 18:48
  • $\begingroup$ Okay, so you're making the transitions to the middle third of the line segment smooth and the transitions from the bottom and top thirds to $f_0$ smooth. That seems like it should work. I;m going to think about it a little more. $\endgroup$ – Ashley Aug 8 '14 at 19:03

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.