diff --git a/Running Start/CSB320 - Machine Learning Concepts/Class 5-21 (Ensemble Models).md b/Running Start/CSB320 - Machine Learning Concepts/Class 5-21 (Ensemble Models).md
index 29095f8..0060aae 100644
--- a/Running Start/CSB320 - Machine Learning Concepts/Class 5-21 (Ensemble Models).md	
+++ b/Running Start/CSB320 - Machine Learning Concepts/Class 5-21 (Ensemble Models).md	
@@ -52,4 +52,8 @@ bootstrap_samples = [
 	- ![[GradientDescent.excalidraw]]
 
 ## Zybooks notes
-- 
\ No newline at end of file
+- Boosting
+	- decision stumps are used (decision trees with 1 layer)
+	- an initial model is fit to the data, it will get many things wrong since it is a weak learner
+	- Then, incorrectly predicted instances are weighted higher than correct ones and a new model is trained with this
+	- This keeps happening and final predictions are made by going back and doing a weighted aggregation of all previous models to create a complex decision boundary from the decision stumps
\ No newline at end of file