Merge branch 'feature/add-adaboost-ensemble' of https://github.com/rathideep22/linfa into feature/add-adaboost-ensemble

Author: rathideep22

algorithms/linfa-ensemble/src/adaboost.rs

@@ -259,7 +259,7 @@ where
             // Handle edge cases for weighted error
             if weighted_error <= 0.0 {
                 // Perfect prediction - add model with maximum weight and stop
-                model_weights.push(10.0); // Large weight for perfect classifier
+                model_weights.push(PERFECT_CLASSIFIER_WEIGHT); // Large weight for perfect classifier
                 models.push(model);
                 break;
             }

algorithms/linfa-ensemble/src/adaboost_hyperparams.rs

@@ -20,7 +20,7 @@ use rand::Rng;
 /// * `model_params`: The parameters for the base learner (weak classifier). Typically, shallow
 ///   decision trees (stumps with max_depth=1 or max_depth=2) are used as weak learners.
 ///
-/// * `rng`: Random number generator used for tie-breaking and reproducibility.
+/// * `rng`: Random number generator used for bootstrap sampling and reproducibility.
 #[derive(Clone, Copy, Debug, PartialEq)]
 pub struct AdaBoostValidParams<P, R> {
     /// The maximum number of estimators to train