80. Machine learning for physics

Applies machine learning to classification, regression, surrogate models, anomaly detection, inverse problems, and scientific discovery. Learners guard against leakage, weak validation, hidden bias, and physically impossible predictions.