Powerful Modified-EDF Goodness-of-Fit Tests
具有较高功效的修正经验分布函数拟合优度检验
Modified forms of the Kolmogorov-Smirnov, Cramér-von Mises and Anderson-Darling goodness-of-fit tests have been studied for the situation where parameters have to be estimated. Using Monte Carlo power studies, it is shown that some of the new tests effect improvements over other known tests.