Paper Title: Automobile Service Upselling Feature Extraction using Multiple Feature Importance Techniques
Conference Name: Korean Institute of Communications and Information Sciences (KICS Fall 2023) Conference
Abstract: With advances in information technology and big data, the branch of customer service and retailing also tread the path of automation. This work leveraged a company dataset having customer service requests for vehicle labor services, with attempts to determine the most impactful features in a dataset having the target feature as the “Total Sales” to create a model that tries to improve the upselling of automobile services. Three feature extraction techniques were utilized: random forest, linear regression, and gradient boosting, the features were ranked in accordance with the increase in root-mean-square error indicating which features have the most impact on the target feature. The common features among the three methods are listed as the results.
Status: Accepted
Conference Type: Domestic
Cite: Paul Michael Custodio, Robin Matthew Medina, Judith Nkechinyere Njoku, Cosmas Ifeanyi Nwakanma, Jae-Min Lee, and Dong-Seong Kim, "Automobile Service Upselling Feature Extraction using Multiple Feature Importance Techniques", Korean Institute of Communications and Information Sciences (KICS Fall 2023