Grid search, walk-forward, out-of-sample testing

This lesson is for enrolled learners

The concept in plain words A strategy almost always has parameters you could tune — moving-average lengths, a stop-loss multiplier, an RSI threshold — and it's tempting to simply try many combinations and keep whichever one produced the best historical return. This is a grid search, and done naively on the full dataset, it's also the fastest way to produce……

Enroll to continue Already enrolled? Sign in