Densest subgraph variation

Let $G=(V,E)$ be a graph. Consider an edge weight function $w:E\to {\mathbb{R}}^{+}$ and a vertex cost function $c:V\to {\mathbb{R}}^{+}$.

We are interested in finding $S\subset V$, such that $w(E(S))-c(S)$ is maximized.

This is very close to the densest subgraph problem, as it is basically the Lagrangian relaxation of the problem.

This problem is equivalent to a min-$st$-cut computation on a suitable graph. Indeed, minimizing $w(E(S))-c(S)$ is equivalent to minimizing $c(S)+\frac{1}{2}w(E(S,\bar{S}))+\frac{1}{2}{\sum }_{v\in \bar{S}}\mathrm{deg}(v)$. This can be solved easily by modeling it as a min-$st$-cut.