Analysis and prediction of customer churn using machine learning - a case study in the banking sector.
By: Emmanuel, Sindikubwabo.
Contributor(s): Marcel, Ndengo.
Publisher: Haryana IOSR - International Organization of Scientific Research 2023Edition: Vol.25(4), Jul-Aug.Description: 27-33p.Subject(s): Computer EngineeringOnline resources: Click here In: IOSR Journal of Computer Engineering (IOSR-JCE)Summary: Customer turnover is a global issue that has an impact on the banking business. This study aims to raise knowledge of whether a customer is likely to switch banks depending on requested services. The several ensemble machine learning algorithms namely, Ada-Boost (AD), Random Forest (RF), Light Gradient Boosted Machine (LGB), and CatBoost (CT) combined to make the proposed technique for our research called Super learner. The study compared the super learner that comes from ensembles to the one produced under the combination and configuration of weaker learners machine learning models precisely, Decision Tree (DT), K Nearest Neighbours (KNN), Support Vector Machine (SV M), and Logistic Regression(LR). We used a churn for bank customer data-set from Kaggle. Super Learning algorithms helped us to categorize customers who are likely to change from one bank to another bank and those who are not. Both of the super learners were able to outperform all of the employed machine learning models. The machine learning evaluation measures assisted us in deciding that the super learner produced by ensemble machine learning models, with an accuracy of 87.7%, was the optimal model to utilize in our research using the data set from kaggle, which recorded past customer’s bank information.Item type | Current location | Call number | Status | Date due | Barcode | Item holds |
---|---|---|---|---|---|---|
Articles Abstract Database | School of Engineering & Technology Archieval Section | Not for loan | 2023-1550 |
Customer turnover is a global issue that has an impact on the banking business. This study aims to raise
knowledge of whether a customer is likely to switch banks depending on requested services. The several
ensemble machine learning algorithms namely, Ada-Boost (AD), Random Forest (RF), Light Gradient Boosted
Machine (LGB), and CatBoost (CT) combined to make the proposed technique for our research called Super
learner. The study compared the super learner that comes from ensembles to the one produced under the
combination and configuration of weaker learners machine learning models precisely, Decision Tree (DT), K
Nearest Neighbours (KNN), Support Vector Machine (SV M), and Logistic Regression(LR). We used a churn for
bank customer data-set from Kaggle. Super Learning algorithms helped us to categorize customers who are
likely to change from one bank to another bank and those who are not. Both of the super learners were able to
outperform all of the employed machine learning models. The machine learning evaluation measures assisted
us in deciding that the super learner produced by ensemble machine learning models, with an accuracy of
87.7%, was the optimal model to utilize in our research using the data set from kaggle, which recorded past
customer’s bank information.
There are no comments for this item.