Table 11 Statistical hypothesis testing results for model comparisons.
From: Intelligent information management enables quality-by-design in pharmaceutical production
Comparison | Metric Tested | Test Applied | Mean Difference | 95% CI | p-value | Significance |
|---|---|---|---|---|---|---|
CNN vs. DoE | R² | Paired t-test | + 0.11 | [0.07, 0.15] | < 0.001 | Significant |
DNN vs. Regression | RMSE | Wilcoxon signed-rank | -0.043 | [-0.06, -0.02] | < 0.01 | Significant |
RF vs. DoE | R² | ANOVA (post-hoc Tukey) | + 0.03 | [-0.01, + 0.07] | 0.09 | Not Significant |
GBM vs. Regression | MAE | Paired t-test | -0.018 | [-0.03, -0.01] | 0.03 | Significant |
CNN vs. RF | F1-score | ANOVA (post-hoc Tukey) | + 0.07 | [0.04, 0.10] | < 0.01 | Significant |
DNN vs. SVM | R² | Wilcoxon signed-rank | + 0.09 | [0.05, 0.13] | < 0.001 | Significant |
