• V. Umayaparvathi
• Prof. K. Iyakutti

Customer churn prediction is the process of identifying the possible churners in advance before they leave the network. It helps the Customer Relationship Management (CRM) department prevent the subscribers who are likely to churn in the future by taking the required retention policies to attract the likely churners and retain them. Machine learning techniques are widely applied to the churn prediction problem, and recent studies compare their performances. In recent years, ensemble learning techniques gained much attention due to their superior performance than traditional algorithms. A voting ensemble involves combining multiple models' predictions to output the final class, e.g., churner or not. This work presents an improved weighted soft-voting ensemble classifier. The weights are assigned to the heterogeneous ensemble members based on their performance during the training phase. The proposed ensemble classifier achieved better accuracy when compared with ten standard machine learning and ensemble models.

Customer relationship management, Churn prediction, Machine learning, and Ensemble learning,

"Customer Churn Prediction using Weighted Soft-voting Ensemble Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 4, page no.1116-1121, April-2019, Available :http://www.jetir.org/papers/JETIR1904U41.pdf

"Customer Churn Prediction using Weighted Soft-voting Ensemble Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 4, page no. pp1116-1121, April-2019, Available at : http://www.jetir.org/papers/JETIR1904U41.pdf