neural_trees.paired_t_test

neural_trees.paired_t_test(clf_A: Any, clf_B: Any, X, y, k_folds: int = 10, alpha: float = 0.05, random_state: int = 42) → TestResult

Paired t-test for classifier comparison.

Splits data into k folds, trains/tests both classifiers on each fold, and applies a paired t-test on the k accuracy differences.

Note: This test has inflated Type I error due to the non-independence of overlapping training sets. Prefer the 5×2cv F test (Alpaydın, 1999).

Parameters:

clf_A, clf_Bsklearn-compatible classifiers

Xarray-like

yarray-like

k_foldsint, default=10

alphafloat, default=0.05

random_stateint, default=42

Returns:

TestResult