-
No. 09Nov 2026
- Prediction and forecasting
- Grinding and classification
- Classical machine learning
- Open and synthetic datasets
A model predicts grinding throughput with R² = 0.962. What did it actually learn?
Every one of the dataset's 75,000 numbers, throughput included, comes out of a random-number generator seeded with 42. The regression learned the straight-line part of the throughput formula, not a mill.