typos

Author: glemaitre

examples/applications/plot_impact_imbalanced_classes.py

@@ -396,8 +396,9 @@ def evaluate_classifier(clf, df_scores, clf_name=None):
 df_scores
 
 ###############################################################################
-# When we analyse the results, we can draw a similar conclusion than in the
+# When we analyse the results, we can draw similar conclusions than in the
 # previous discussion. However, we can observe that the strategy
-# `class_weight="balanced"` does not improve the performance. A resampling is
-# indeed required. The most effective method remains the
-# `BalancedBaggingClassifier` using a GBDT as a base learner.
+# `class_weight="balanced"` does not improve the performance when using a
+# `RandomForestClassifier`. A resampling is indeed required. The most effective
+# method remains the `BalancedBaggingClassifier` using a GBDT as a base
+# learner.