Volume 11 Issue 2 - August 2018

  • 1. Performance analysis of machine learning algorithms in customer churn prediction

    Authors : Deepthi Das, Dr. Raju Ramakrishna Gondkar

    Pages : 29-34

    DOI : http://dx.doi.org/10.21172/1.112.06

    Keywords : Customer attrition Customer Relationship Management (CRM) Support Vector Machines (SVM) Artificial Neural Network (ANN)

    Abstract :

    Customer attrition is termed by several industrialists and e-commerce professionals to recognize the customers, who are about to change their service from the existing company or end their period of subscription. In recent years, companies such as e-commerce, telecommunication and insurance sectors are facing tremendous pressure due to financial disintermediation and marketing and the gradual increase in the competitiveness tends to provide better service with lesser cost. So, early prediction of the behaviour of the clients plays an important role in the real-time market and can help to retain the loyal customers. In this research, a survey on different data mining techniques and machine learning algorithms along with the challenges of customer attrition prediction in the motor insurance sector are depicted. The survey on the application of the various machine learning algorithm for churn prediction is mainly observed in telecommunication sector and Support Vector Machine (SVM), Artificial Neural Network (ANN) are generally used algorithm for churn analysis and forecasting. Various authors have considered different tools for analysis and the result obtained from the study shows that combination of the two-step process of ANN for training and combined approach of SVM for testing provides better accuracy with high Area Under the Curve compared to existing techniques.

    Citing this Journal Article :

    Deepthi Das, Dr. Raju Ramakrishna Gondkar, "Performance analysis of machine learning algorithms in customer churn prediction", Volume 11 Issue 2 - August 2018, 29-34