Table 17 Model comparison criteria: Mean BIC, mean DIC, mean ML, and mean fraction of non-rejected normality tests with KS for the unperturbed Ehrlich-tumor data.
From: Proper likelihood functions for parameter estimation in S-shaped models of unperturbed tumor growth
Model | \({\overline{BIC}}\) | \({\overline{DIC}}\) | \({\overline{ML}}\) | \(\overline{fr(KS)}\) |
|---|---|---|---|---|
Norm | \(9.7976 \times 10^{1}\) | \(1.4196 \times 10^{2}\) | \(\mathbf {9.7207 \times 10^{-1}}\) | \(2.0000 \times 10^{-4}\) |
Prop | \(1.0433 \times 10^{2}\) | \(1.4442 \times 10^{2}\) | \(9.2043 \times 10^{-1}\) | \(2.7164 \times 10^{-1}\) |
Thres | \(9.6017 \times 10^{1}\) | \(\mathbf {1.3925 \times 10^{2}}\) | \(8.0212 \times 10^{-1}\) | \(\mathbf {2.7512 \times 10^{-1}}\) |
Stud | \(9.7980 \times 10^{1}\) | \(1.4328 \times 10^{2}\) | \(9.6507 \times 10^{-1}\) | \(2.4000 \times 10^{-4}\) |
StudProp | \(\mathbf {9.2215 \times 10^{1}}\) | \(1.4498 \times 10^{2}\) | \(5.6668 \times 10^{-1}\) | \(2.7290 \times 10^{-1}\) |
