Sigma-subdifferential and its application to minimization problem (Q2188356)

The authors study a generalized class of convex functions, that they call \(\sigma \)-convex, consisting of functions \(f:X\rightarrow \mathbb{R}\cup \{+\infty \}\) that satisfy \[ f(tx+(1-t)y)\leq tf(x)+(1-t)f(y)+\sigma (x)t(1-t)||x-y||, \] for all \(x,y\in X\) and \(t\in \lbrack 0,1]\), where \(\sigma :X\rightarrow \mathbb{R}_{+}\cup \{+\infty \}\) is any positive-valued function whose domain contains the domain of \(f\). They consider an adapted notion of subdifferential for these functions and reproduce results -- knowing to hold for convex functions -- for this new class, including first-order optimality conditions.