Data-Driven Retention Strategies: Exploring the Efficacy of Composite Deep Learning for Customer Churn Prediction in Telecommunications

The customer churn, otherwise known as subscriber loss, inclusive among one of the key challenges of the telecommunication industry (TCI). Forecasting churn gives rise to telecoms step-saving retentional strategies, serving as the means preventing losing revenue. To this research applies of the innovative customer churn prediction (CCP) model employing data transformation approaches and feature selection schemes. The work entail analyzing six data transformation methods (Log, Rank, Box-Cox, Z-score, Discretization, and weight-of-evidence) and eight classifiers such that each data conversion technique is applied or tested on multiple types of classifiers such as K-Nearest Neighbors (KNN), Naive Bayes (NB), Logistic Regression (LR), Random Forest (RF) The univariate way of feature selection is used to determine the features which have major effect and the proposal grid search is also adopted for hyperparameter optimisation. Experiments are conducted on four selected telecommunication industry (TCI) datasets. The evaluation of the model's performance involves using metrics such as accuracy and recall which are understood to be better than existing tools.