Penerapan Algoritma XGBoost Dalam Menganalisa Keberlanjutan Pelanggan Tour dan Travel

Abstract

Customer churn is a term used to describe customer loss in the business world.
Customer churn is a major challenge in the business world, impacting every company. One example is the tour and travel industry, the impact of customer churn in tour and travel businesses can include decreased profits and increased operational costs because acquiring tour and travel customers is more expensive than retaining them. Every company has a strategy for customer retention, one of which is implementing machine learning. This study uses public data to determine customer churn using the XGBoost algorithm. Extreme Gradient Boost (XGBoost) works by gradually building a model to improve prediction accuracy. In this study, the XGBoost model works through several stages: data processing, dataset division, algorithm testing, which ultimately results in model accuracy, evaluation, and ROC and AUC curves. The results of this study with the XGBoost model produced an accuracy of 87.7%, precision of 74.4%, recall of 74.4%, F1-Score of 74.4%, and an AUC value of 0.95. In addition, this study also produced an application to predict customers who do not yet have a label