DefinePK

Abstract

Despite the significant flow of remittances into Uzbekistan, the adoption rate of financial services and products among remittance recipients remains relatively low. Identifying which remittance recipients are more suitable for cross-selling retail banking products is crucial. The primary objective of this study is developing a robust model that assists financial institutions to identify potential customers with a higher likelihood of adopting car loans. We use a unique dataset of remittance transactions and vehicle financing data provided by a commercial bank in Uzbekistan. To balance data, we apply Synthetic Minority Over-sampling Technique for Nominal and Continuous variables (SMOTE-NC) and combination of over-sampling approach of SMOTE with an under-sampling method called Edited Nearest Neighbors (SMOTE-ENN). We compare the performance of Logistic Regression (LR), Decision Tree (DT), Support Vector Machine (SVM), and Random Forest (RF) models in these two datasets. Our analysis reveals than all models perform better on the SMOTEEN dataset. DT and RF outperform the other models. Based on these insights, we recommend using DT, RF and SMOTEEN techniques on imbalanced datasets. The results of this study offer practical implications for data scientists and financial institutions in remittance-receiving countries aiming to leverage remittance flows and boost cross-selling.
Keywords: Cross-Selling, Machine-Learning, Remittances, Vehicle Loans, Uzbekistan.