TY - GEN
T1 - Machine Learning Techniques for Predicting Customer Churn in A Credit Card Company
AU - Chang, Victor
AU - Gao, Xianghua
AU - Hall, Karl
AU - Uchenna, Emmanuel
N1 - Copyright ©2022 IEEE
PY - 2023/3/23
Y1 - 2023/3/23
N2 - As marketplaces have become increasingly crowded, businesses have recognized the importance of focusing their business strategy on identifying customers who are likely to leave their services. To solve this, a technique for identifying these consumers must launch pre-emptive retention efforts to keep them. Therefore, to minimize costs and maximize efficiency, churn prediction must be as precise as possible to guarantee retention efforts are directed solely at customers who intend to transfer service providers. The study conducted in this report aims to establish a mechanism for anticipating churn in advance while minimizing misclassification. The suggested methodology integrates a temporal dimension into customer churn prediction to maximize future attrition capture by identifying probable customer loss as soon as possible. Six machine learning algorithms are selected and conducted to validate the suggested methodology using a bank credit card dataset. Finally, the proposed methodology’s results are compared to those published churn prediction methodologies. According to the research, clients can be classified into clusters based on their contracts with the service provider. It is conceivable to estimate when the customer might be expected to end their service with the organization.
AB - As marketplaces have become increasingly crowded, businesses have recognized the importance of focusing their business strategy on identifying customers who are likely to leave their services. To solve this, a technique for identifying these consumers must launch pre-emptive retention efforts to keep them. Therefore, to minimize costs and maximize efficiency, churn prediction must be as precise as possible to guarantee retention efforts are directed solely at customers who intend to transfer service providers. The study conducted in this report aims to establish a mechanism for anticipating churn in advance while minimizing misclassification. The suggested methodology integrates a temporal dimension into customer churn prediction to maximize future attrition capture by identifying probable customer loss as soon as possible. Six machine learning algorithms are selected and conducted to validate the suggested methodology using a bank credit card dataset. Finally, the proposed methodology’s results are compared to those published churn prediction methodologies. According to the research, clients can be classified into clusters based on their contracts with the service provider. It is conceivable to estimate when the customer might be expected to end their service with the organization.
KW - credit card churn prediction
KW - exploratory data analysis
KW - machine learning
UR - https://ieeexplore.ieee.org/document/10077145
U2 - 10.1109/IIoTBDSC57192.2022.00045
DO - 10.1109/IIoTBDSC57192.2022.00045
M3 - Conference publication
T3 - Proceedings - 2022 International Conference on Industrial IoT, Big Data and Supply Chain, IIoTBDSC 2022
SP - 199
EP - 207
BT - 2022 International Conference on Industrial IoT, Big Data and Supply Chain (IIoTBDSC)
PB - IEEE
T2 - 2022 International Conference on Industrial IoT, Big Data and Supply Chain (IIoTBDSC)
Y2 - 23 September 2022 through 25 September 2022
ER -