# ================================================================================
# ================================= CORE METRICS =================================
# ================================================================================
===== FINAL SUMMARY =====
Best epoch : 128
Train accuracy : 0.988000
Val accuracy : 0.985333
Train loss : 0.012553
Val loss : 0.014648
Threshold : 0.520000
Test accuracy : 0.988000
Test loss : 0.051786 ===== TRAIN =====
Classification Report| precision | recall | f1-score | support | |
|---|---|---|---|---|
| 0 | 0.991773 | 0.999585 | 0.995664 | 2412.000000 |
| 1 | 0.999065 | 0.981618 | 0.990264 | 1088.000000 |
| accuracy | 0.994000 | 0.994000 | 0.994000 | 3500.000000 |
| macro avg | 0.995419 | 0.990602 | 0.992964 | 3500.000000 |
| weighted avg | 0.994040 | 0.994000 | 0.993985 | 3500.000000 |
| Positive (1) | Negative (0) | |
|---|---|---|
| Positive (1) | 1068 | 1 |
| Negative (0) | 20 | 2411 |
AUC/AUPRC AUC (ROC): 0.999775 AUPRC: 0.999513
===== VALIDATION =====
Classification Report| precision | recall | f1-score | support | |
|---|---|---|---|---|
| 0 | 0.983773 | 0.989796 | 0.986775 | 490.000000 |
| 1 | 0.980545 | 0.969231 | 0.974855 | 260.000000 |
| accuracy | 0.982667 | 0.982667 | 0.982667 | 750.000000 |
| macro avg | 0.982159 | 0.979513 | 0.980815 | 750.000000 |
| weighted avg | 0.982654 | 0.982667 | 0.982643 | 750.000000 |
| Positive (1) | Negative (0) | |
|---|---|---|
| Positive (1) | 252 | 5 |
| Negative (0) | 8 | 485 |
AUC/AUPRC AUC (ROC): 0.994168 AUPRC: 0.975018
===== TEST =====
Classification Report| precision | recall | f1-score | support | |
|---|---|---|---|---|
| 0 | 0.990584 | 0.992453 | 0.991517 | 530.000000 |
| 1 | 0.981735 | 0.977273 | 0.979499 | 220.000000 |
| accuracy | 0.988000 | 0.988000 | 0.988000 | 750.000000 |
| macro avg | 0.986159 | 0.984863 | 0.985508 | 750.000000 |
| weighted avg | 0.987988 | 0.988000 | 0.987992 | 750.000000 |
| Positive (1) | Negative (0) | |
|---|---|---|
| Positive (1) | 215 | 4 |
| Negative (0) | 5 | 526 |
AUC/AUPRC AUC (ROC): 0.998306 AUPRC: 0.996166
Scenario D emits three structured log tables that document ensemble behavior and make the MAIN vs TEMP workflow auditable and reproducible.
Main Log (main_log) —
Iteration-level snapshots of the Primary (MAIN) ensemble state and
evaluation results under the selected metric.
Movement Log (movement_log) —
Deterministic promotion and replacement events between TEMP and MAIN
(what moved, directionality, and why).
Change Log (change_log) —
Per-iteration update diagnostics and structural deltas recorded during
training and selection steps.
These tables are returned in res_D$runs[[1]]$tables.
The previews below are capped for vignette readability.
| serial | iteration | phase | metric_name | metric_value | message | timestamp |
|---|---|---|---|---|---|---|
| 0.0.1 | 1 | main_before | accuracy | 0.9786667 | 2026-03-08 22:57:12 | |
| 0.0.2 | 1 | main_before | accuracy | 0.9840000 | 2026-03-08 22:57:12 | |
| 0.0.1 | 1 | main_after | accuracy | 0.9826667 | 2026-03-08 22:59:31 | |
| 0.0.2 | 1 | main_after | accuracy | 0.9760000 | 2026-03-08 22:59:31 | |
| 0.0.1 | 2 | main_before | accuracy | 0.9826667 | 2026-03-08 22:59:31 | |
| 0.0.2 | 2 | main_before | accuracy | 0.9760000 | 2026-03-08 22:59:31 | |
| 0.0.1 | 2 | main_after | accuracy | 0.9826667 | 2026-03-08 23:01:47 | |
| 0.0.2 | 2 | main_after | accuracy | 0.9760000 | 2026-03-08 23:01:47 |
| serial | iteration | message | timestamp |
|---|---|---|---|
| 0.0.2 | 1 | removed (no replacement) | 2026-03-08 22:59:31 |
| 0.0.2 | 2 | removed (no replacement) | 2026-03-08 23:01:47 |
| serial | iteration | message | timestamp |
|---|---|---|---|
| 0.0.2 | 1 | model removed from main | 2026-03-08 22:59:31 |
| 0.0.2 | 2 | model removed from main | 2026-03-08 23:01:47 |
Note: Tables below are preview-capped for vignette
readability. Full tables remain available in res_D\(runs[[1]]\)tables. Artifact writing
is OFF by default for CRAN-safety.