THE RESULT
What the data shows
The largest computed metric is 6,281 for K_Scatch; the smallest is 16.5 for Stains. Metric: median_pixels (Pixels).
Image-scale and acquisition conditions may differ; geometry is not a physical damage severity measure.
| fault | records | median pixels | median x perimeter | median y perimeter |
|---|---|---|---|---|
| Bumps | 402 | 120.5 | 17 | 16 |
| Dirtiness | 55 | 145 | 28 | 35 |
| K_Scatch | 391 | 6,281 | 271 | 139 |
| Other_Faults | 673 | 146 | 25 | 22 |
| Pastry | 158 | 209 | 20 | 30.5 |
| Stains | 72 | 16.5 | 7 | 4.5 |
| Z_Scratch | 190 | 146 | 28 | 23 |
THE METHOD
From source to answer
Class-level median geometry with sample sizes.
The Python source exposes this study’s transformations. The complete project download includes shared preparation and evaluation routines.
THE NEXT DECISION
What follows from the finding
Discuss whether geometry can support inspection triage with process engineers.
Where the conclusion stops
Every source row is a recorded fault. The data cannot estimate a production defect rate or distinguish healthy plates. Batch and machine IDs are absent, so grouped exact-input holdouts do not prove factory transfer.
Related studies may reuse observations or holdouts. These are historical analyses; associations and backtests do not demonstrate commercial impact. Further model tuning needs new, untouched evaluation data.