Skills - Python, Machine Learning Algos (Naive Bayes, Decision Tree, Adaboost)
-
Analysis of various customer-related data such as transaction history, demographics, account activity, and customer interactions, the model aims to predict which customers are most likely to churn in the near future.
-
In Business Context, the percentage of FN is important. A FN predicts a customer will not churn when they actually do. So, FN should be minimized.
-
The best model selected is AdaBoost with an average recall-FN for the testing data of 0.998333. 👍