diff --git a/14_imbalanced/handling_imbalanced_data_exercise.md b/14_imbalanced/handling_imbalanced_data_exercise.md
index 8aa2cea..c86c30b 100644
--- a/14_imbalanced/handling_imbalanced_data_exercise.md
+++ b/14_imbalanced/handling_imbalanced_data_exercise.md
@@ -14,6 +14,17 @@
     1. Improve f1 score in minority class using various techniques such as undersampling, oversampling, ensemble etc
     
     [Solution](https://github.com/codebasics/deep-learning-keras-tf-tutorial/blob/master/14_imbalanced/Handling%20Imbalanced%20Data%20In%20Customer%20Churn%20Using%20ANN/Bank%20Turnover%20Customer%20Churn%20Using%20ANN.ipynb)
-    
-    Thanks https://github.com/src-sohail for providing this solution.
+   Thanks https://github.com/src-sohail for providing this solution.
+   3. Exercise: Predicting Customer Satisfaction
+   Use the Customer Satisfaction dataset from Kaggle. - https://www.kaggle.com/datasets/teejmahal20/airline-passenger-satisfaction
+
+      1. Build a classification model to predict customer satisfaction.
+      2. Initially, use a logistic regression model from scikit-learn.
+      3. Print the classification report and analyze precision, recall, and f1-score.
+      4. Try to improve the f1-score for the minority class using techniques like undersampling, oversampling, or ensemble methods.
+
+      5. [Solution] : https://www.kaggle.com/code/teejmahal20/classification-predicting-customer-satisfaction 
+   
+   Thanks https://kaggle/teejmahal20 for providing this solution.
+