Asian Journal of Basic Science & Research Volume 4, Issue 4, Pages 60-73, October-December 2022 ISSN: 2582-5267 60 Customer Churn Prediction Using Machine Learning Techniques: the case of Lion Insurance Edemealem Desalegn Kingawa 1,2,* & Tulu Tilahun Hailu 1,2 1 Artificial Intelligence and Robotics Center of Excellence, Addis Ababa Science and Technology University, Addis Ababa, Ethiopia. 2 Department of Software Engineering, College of Electrical and Mechanical Engineering, Addis Ababa Science and Technology University, Addis Ababa, Ethiopia. Corresponding Author – (Edemealem Desalegn Kingawa): desuking1717@gmail.com* DOI: http://doi.org/10.38177/AJBSR.2022.4407 Copyright: © 2022 Edemealem Desalegn Kingawa & Tulu Tilahun Hailu. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Article Received: 15 November 2022 Article Accepted: 18 December 2022 Article Published: 26 December 2022 ░ 1. INTRODUCTION Insurance is a crucial part of financial planning. The insurance industry is a large institution that protects the assets of organizations or individuals. Therefore, the insurance company is set up to provide important services to its customers. This depends on the premium paid by the customers [1]. So insurance is a way to manage risk. The insurance sector has become one of the main industries in both developed and developing countries. Modern Insurance was started in Great London when a Fire accident happened in 1666, after destroying more than 30,000 homes; a man named Nicholas Barban began building insurance for the first time. He later introduced the city's first fire insurance corporation. Accident insurance was introduced in the late 19th century and was very similar to modern disability coverage at that time [2]. After 355 years, this type of service has become mandatory by the government. This is because of the enormous benefits that insurance offers to human beings. Except for Ethiopia and Liberia, most of the African countries have been colonized by other countries, especially Europe. It was a frustrating time for Africa, but it also brought new ideas to Africa. As a result, these colonial powers expanded infrastructures like railways, ports roads, and airports to facilitate their power. This new idea enables them to build and expand infrastructures to maximize their benefits. These highly invested infrastructures have brought to Africa a variety of policies and strategies that can be used by insurance companies to prevent risk [3]. Insurance has been a major competitor sector in Ethiopia and around the world. It was established in our country in 1984 G.C its name is known as Ethiopian Insurance Corporation (EIC). Currently, our country has grown to 14 insurance service providers in 37 years [4]. The increasing number of insurance providers raised the level of ABSTRACT The growth of an insurance company is measured by the number of policies purchased by customers. To keep the company growing and having more customers, the customer churn prediction model is crucial to maintain its competitiveness. Even if the company has good service delivery, it is important to identify the customer’s behavior and be able to predict the future churners. The mai n contribution to our work is the development of a predictive model that can proactively predict the customer who will leave the insurance company. The model developed in this study uses machine learning techniques on lion insurance data. Another main contribution of this study is the labeling of the data using an unsupervised algorithm on 12007 rows with 9 features from which 2 clusters were generated using the K-means++ algorithm. As the cluster results found are imbalanced, the synthetic minority oversampling technique was applied to the training dataset. The Deep Neural Network algorithm turns out to be a very effective model for predicting customer churn, reaching an accuracy of 98.81%. The two years of customer data were obtained from lion insurance and used to train test, and evaluate the model. The Randomized optimization technique was selected for each algorithm. However, the best results were obtained by a deep neural network with a structure of (9-55-55-55-55-55-1). This algorithm was selected for classification in this churn prediction study. Keywords: Churn prediction; Clustering; Supervised machine learning; Deep neural learning; Sampling; Encoding.