Prediction of Likely Customers for Car Industries Using K-Means Clustering Compared with Logistic Regression

Abstract

This study represents a customer segmentation model which helps them to group the customers for the car industry with the same market characteristics. The study contains 2 groups i.e, the Logistic Regression model is developed in the first group and the K Means clustering model, an unsupervised machine learning algorithm is developed in the second group. Each group has a sample size of 200 and the study parameters include alpha value0.05, beta value 0.2, and the power value 0.8. The accuracies of each model are compared with others for different sample sizes. This article is an attempt to improve the accuracy of customer segmentation using the K Means clustering, an unsupervised clustering machine learning algorithm. The proposed model has improved accuracy of 87.4% with p < 0.05 in segmenting customers than the existing model of 85%. This innovative prediction model helps to know the future customers and their needs and take innovative decisions to fulfill them. The outcomes of the proposed model are compared with the Logistic Regression algorithm and the proposed model confirms to have higher accuracy than the Logistic Regression algorithm.