← Research logMachine Learning
KILLED · overfit stack
#

Stacking many weak predictors under a meta-learner yields a strong combined return forecast.

$$ \hat y = g\big(f_1(\mathbf{x}),f_2(\mathbf{x}),\dots,f_m(\mathbf{x})\big) $$

Stack LSTM, XGBoost, logistic, GP base learners under a meta-regressor with nested cross-validation.

Cross-val improvementmarginal
Out-of-sample (true holdout)worse than best single
Leakage via the stackhard to fully avoid
KILLED
Stacking amplifies overfitting and leakage when the base learners share near-zero-signal targets — the meta-learner fits the validation noise. On a true holdout it underperformed the single best model. Complexity tax with negative return. Killed.
Ensembling weak-or-zero-signal models mostly ensembles their overfitting. Stacking is not a substitute for a target that has signal.

We publish the failures too.

This is one of 100+ documented hypotheses. Browse the full lab notebook, or see the strategies that survived into production.