Analysis of Customer Churn for Telecom Company with SMOTE-ENN and Hyperparameter Tuning Randomized-SearchCV Technique in Advanced Machine Learning Technology

Customer churn analysis is an important task for telecom companies which aims to retain their customer base and enhance profitability. This paper proposes to help telecom companies predict the potential churn rate of their customers by using advanced machine learning technology. Specifically, this research intends to build a model with a Balanced Dataset and performance hyperparameter tuning using Randomized SearchCV. In this paper, the unbalance in the training set is subsequently addressed using the Synthetic Minority Oversampling Technique (SMOTE) with Edited Nearest Neighbor (ENN) and increases the performance of the model using Hyperparameter Tuning Randomized SearchCV. Following that, the effectiveness of advanced machine learning methods is compared by using "With Hyperparameter Tuning" and "Without Hyperparameter Tuning" techniques. Finally, this research focuses on providing valuable insights for telecom companies to proactively retain customers and enhance the overall satisfaction of customers.