Klasifikasi Customer Relationship Management Perusahaan Telekomunikasi Seluler Dengan Metode Machine Learning

Abstract

Customer Relationship Management (CRM) as the part of Enterprise Resource Planning (ERP) focuses on the company relationship with the costumer must be optimized. By the increasingly of tight business competition, CRM the cellular telecommunication company must be able to carry out the classification of loyal costumer and moving costumer (churn) as the initial steps to maintain costumer. This research aims to classification the numbers of costumers who are churn from the company to another company. The research approach was conducted by quantitative methods using secondary data. The data research was obtained from the data source of telco_dataset.csv. The data input was conducted by Python software. The results of the data collection are processed through the machine learning by decision tree, logistic regression and random forest methods. The results of research show the classifications by the methods of decision tree 78, 27%, logistic regression 78, 67% and random forest 78, 13% has the accuracy. Then, the classification model that suitable can be used to classify Customer Relationship Management. The cellular telecommunication company is the logistic regression method, because has more the high accuracy level. As the factors that have sensitive influential in classification of establishment that are the total of cost per month are issued by customers, the total of overall cost during being customers and the how long customers subscribe is.