 Squared-error loss

Quadratic loss penalizes an estimate $a$ of a vector parameter $\theta$ by $\Vert a-\theta {\Vert }^2$. Its posterior expected value is minimized by the posterior mean.