sb_bagging

Sampling, Validation and ML Diagnostics

Sequentially bootstrapped bagging classifiers/regressors.

Why This Module Exists

Combines ensemble variance reduction with overlap-aware sampling.

Key Public APIs

SequentiallyBootstrappedBaggingClassifier
SequentiallyBootstrappedBaggingRegressor
MaxSamples
MaxFeatures

Core Math

Bagging Predictor

\[\hat f(x)=\frac{1}{B}\sum_{b=1}^{B} f_b(x)\]

Bootstrap Sampling

\[S_b\sim P_{seq}(u)\]

Code Examples

Instantiate SB bagging classifier

use openquant::sb_bagging::SequentiallyBootstrappedBaggingClassifier;

let bag = SequentiallyBootstrappedBaggingClassifier::new(100);

Implementation Notes

Sequential bootstrap improves diversity under event overlap.
Tune max_samples/max_features with out-of-sample monitoring.