More about Discovery Process Models
75
prediction interval was used to measure the performance ability of
LDSCV and NDSCV (i.e., if the interval could include the population value, then the predictions were acceptable). LDSCV predicts all
the population values within the interval except the Weibull population when n = 30. As n increases to 50, the LDSCV intervals include all
population values.
Figure 4.8. (A–D) Pool-size-by-rank plots for a lognormal population derived
by LDSCV when n = 30 (A) and n = 50 (B), and plots derived by NDSCV when
n = 30 (C) and n = 50 (D). The prediction interval is the 0.9 probability level.
75
prediction interval was used to measure the performance ability of
LDSCV and NDSCV (i.e., if the interval could include the population value, then the predictions were acceptable). LDSCV predicts all
the population values within the interval except the Weibull population when n = 30. As n increases to 50, the LDSCV intervals include all
population values.
Figure 4.8. (A–D) Pool-size-by-rank plots for a lognormal population derived
by LDSCV when n = 30 (A) and n = 50 (B), and plots derived by NDSCV when
n = 30 (C) and n = 50 (D). The prediction interval is the 0.9 probability level.