Consider a loss function $h(x,\xi)$ which depends on the decision variables $x$ and uncertain parameters $\xi$. The goal is to minimize the <i>expected value</i> of ...