I find that the following two types of functions are useful to my research.

(i) We know that a function $f: \mathbb{R}_+^m\rightarrow \mathbb{R}$ is called convex if for all ${\bf x,y}\in \mathbb{R}_+^m$ and all $\lambda_1,\lambda_2\geq 0$ s.t. $\lambda_1+\lambda_2=1$, it holds that \begin{equation}f(\lambda_1{\bf x}+\lambda_2 {\bf y})\leq \lambda_1f({\bf x})+\lambda_2f({\bf y}).\end{equation} The first type of function is to drop the restriction of $\lambda_1+\lambda_2=1$, i.e. to replace convex combination with conic combination.

(ii) The second type of function: $f({\bf 0})=0$, and for all ${\bf x},{\bf y},{\bf z}\in \mathbb{R}_+^m$ such that ${\bf y}\geq {\bf x}$, it holds that \begin{equation}f({\bf y})-f({\bf x})\leq f({\bf y}+{\bf z})-f({\bf x}+{\bf z}).\end{equation}

Could anyone tell me whether they are new or not? Are there any useful references? Many thanks!

  1. If you throw in positive homogeneity, then the first class of functions is what is called sublinear, see for instance Proposition 1.1.4 ("Fundamentals of Convex Analysis"; Hiriart-Urruty, Claude Lemaréchal).

  2. The second class of functions that you mention is essentially the same as the Wright convex functions, see for instance pg. 223 of A.W. Roberts, D.E. Varberg, "Convex functions", Academic Press, 1973.

| cite | improve this answer | |
  • $\begingroup$ Thank you very much for the quick and extremely helpful answer! $\endgroup$ – Zhigang Cao Nov 7 '15 at 5:04

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.