ml_paramscan Stable-plateau parameter selection

The overfit cure: when you scan a parameter, don’t pick the single BEST-scoring value (that cell is almost always overfit) — pick the robust PLATEAU that scores well AND consistently. Score each variant by mean − λ·std and choose the argmax. Turns any parametric script self-tuning and overfit-resistant.

Design: the host runs the variants in parallel (a small state array per candidate parameter) and keeps each variant’s running performance mean and std; this library scores and selects. Pure functions of the arrays you pass; nothing repaints.

Concept credit: stability-over-optimum (plateau) parameter selection is a standard walk-forward / robustness practice — original, dependency-free Pine v6 packaging. Educational; not financial advice.

Library "ml_paramscan"

plateauScore(mean, std, lambda)   Parameters:     mean (float)     std (float)     lambda (simple float)

stablePlateau(means, stds, lambda)   Parameters:     means (array<float>)     stds (array<float>)     lambda (simple float)

bestPlateauScore(means, stds, lambda)   Parameters:     means (array<float>)     stds (array<float>)     lambda (simple float)