Sign up ×
MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

I would like to have an inverse (or/and) implicite function theorem for DC-functions. It seems that I have right definitions, but I fail to prove it...


  • Let $h:\mathbb R^n\to\mathbb R$ be a convex function and $x\in\mathbb R^n$, the set of all linear functionals $\ell:\mathbb R^n\to\mathbb R$ such that $$h(y)\ge h(x)+\ell(y-x)$$ is called subdifferential of $h$ at $x$ --- it will be denoted as $\partial_{x}h$. (In general $\partial_{x}h$ is a nonempty bounded convex set)

  • $f:\mathbb R^n\to\mathbb R$ is called DC-function if it is a difference between two convex functions.

  • $F:\mathbb R^n\to\mathbb R^k$ is called DC-map each coordinate function of $F$ is DC.

  • $x\in\mathbb R^n$ is called regular value of a DC-function $f:\mathbb R^n\to\mathbb R$ if of $f=h_1-h_2$ for some convex functions $h_1$ and $h_2$ and $\partial_x h_1 + (-\partial_x h_2)\not\ni0$. Here $+$ denotes Minkowski sum and $(-\partial_x h_2)$ is reflection of $\partial_x h_2$ in the origin.

  • $x\in\mathbb R^n$ is called regular value of a DC-map $f:\mathbb R^n\to\mathbb R^k$ if $x$ is a regular value of $\ell\circ F$ for any non-zero linear map $\ell:\mathbb R^k\to\mathbb R$.

share|cite|improve this question

1 Answer 1

According to this paper an implicit function theorem for delta-convex functions is given in Theorem 4.4 of L. Veselý and L. Zajíček, "Delta-convex mappings between Banach spaces and applications." Dissertationes Math. (Rozprawy Mat.) 289 (1989).

share|cite|improve this answer
FYI, if you can't get a copy of this easily, Zajíček's web page is – Steve Huntsman Jun 8 '10 at 23:07
This theorem only says that implicit function is DC if it exists... Just in case, here is a link to the paper: – Anton Petrunin Jun 10 '10 at 16:57

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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