An Effective Architectural Model for Early Churn Prediction – NELCO
R. Manivannan1, R. Saminathan2, S. Saravanan3
1R. Manivannan*, Research Scholar, Department of Computer Science and Engineering, Annamalai University, Annamalainagar, India.
2Dr. R. Saminathan, Associate Professor, Department of Computer Science and Engineering, Annamalai University, Annamalainagar, India.
3Dr. S. Saravanan, Assistant Professor, Department of Computer Science and Engineering, Annamalai University, Annamalainagar, India.
Manuscript received on July 20, 2019. | Revised Manuscript received on August 10, 2019. | Manuscript published on August 30, 2019. | PP: 4667-4672 | Volume-8 Issue-6, August 2019. | Retrieval Number: F9147088619/2019©BEIESP | DOI: 10.35940/ijeat.F9147.088619
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Customer is an asset of any business organization, whose probable chances of churn is loss. Several challenges are to be considered towards controlling customer churn. Machine learning approach is needed to predict an early churn. Even though various soft computational approaches had been proposed, an optimized computational approach which identifies early churn prediction is necessary. The proposed approach NELCO predicts early customer churn using Negative Correlation Learning (NCL) which uses k-means neighbourhood discriminant similarity indices over network of ensemble values. NELCO proves to have an optimal accuracy towards early prediction of churn, as well as suggests that customer retention rate is higher over PSO, ACO approaches.
Keywords: Negative Correlation Learning, Early prediction, customer churn, customer retention.