TY - JOUR
T1 - Measuring what's missing
T2 - Practical estimates of coverage for stochastic simulations
AU - Jones, Edward Samuel
PY - 2016/6/12
Y1 - 2016/6/12
N2 - Stochastic sensitivity analyses rarely measure the extent to which realized simulations cover the search space. Rather, simulation lengths are typically chosen according to expert judgement. In response, this paper recommends a novel application of Good-Turing estimators of missing distributional mass. Using the United Nations Development Programme's Human Development Index, the empirical performance of such coverage metrics are compared to alternative measures of convergence. The former are advantageous – they provide probabilistic estimates of simulation coverage and permit calculation of strict bounds on estimates of pairwise dominance (for all possible weight vectors, how often country X dominates country Y).
AB - Stochastic sensitivity analyses rarely measure the extent to which realized simulations cover the search space. Rather, simulation lengths are typically chosen according to expert judgement. In response, this paper recommends a novel application of Good-Turing estimators of missing distributional mass. Using the United Nations Development Programme's Human Development Index, the empirical performance of such coverage metrics are compared to alternative measures of convergence. The former are advantageous – they provide probabilistic estimates of simulation coverage and permit calculation of strict bounds on estimates of pairwise dominance (for all possible weight vectors, how often country X dominates country Y).
KW - Faculty of Social Sciences
KW - sensitivity analysis
KW - uncertainty analysis
KW - Monte Carlo
KW - simulation coverage
KW - HDI
U2 - 10.1080/00949655.2015.1077839
DO - 10.1080/00949655.2015.1077839
M3 - Journal article
SN - 0094-9655
VL - 86
SP - 1660
EP - 1672
JO - Journal of Statistical Computation and Simulation
JF - Journal of Statistical Computation and Simulation
IS - 9
ER -