Apa itu Multilevel Monte Carlo?

Multilevel Monte Carlo (Giles, 2008) writes the finest-level expectation as a telescoping sum $\E P_L = \E P_0 + \sum_{l=1}^L\E[P_l - P_{l-1}]$, where $P_l$ is the payoff computed with step $h_l = 2^{-l}T$, and estimates each term independently, with $P_l$ and $P_{l-1}$ in each correction computed from the same Brownian path.