Multilevel and Quasi Monte Carlo Methods for the Calculation of the Expected Value of Partial Perfect Information