Buy Now, Pay Later: An AI and Machine Learning-Based Analysis of Its Influence on Youth Spending Behaviour and Financial Decision-Making (Zenodo)

Summary

• The study analyses how BNPL services influence spending behaviour and financial decision-making among young consumers aged 18–30 using AI and machine learning.p.1
• Primary data were collected from 110 respondents, of whom 73 reported previous BNPL usage.p.4
• Random Forest achieved the highest mean accuracy of 72.76% and F1-score of 84.12% among the three machine learning models.p.3
• BNPL usage frequency was identified as the most important feature in the Random Forest model.p.3

Key insights

1. Young consumers' BNPL use raises concerns about impulsive purchasing, excessive spending and weak financial decision-making.p.2
2. Existing studies offer limited insight into applying AI and machine learning to BNPL spending behaviour.p.2
3. The Random Forest used an ensemble of 100 decision trees.p.3
4. The Decision Tree used a maximum tree depth of 3.p.3
5. The evaluated algorithms did not demonstrate predictive improvement over the baseline on the available dataset.p.3
6. The survey indicated that a substantial proportion of respondents perceived BNPL as contributing to impulsive purchases and increased spending.p.4
7. The results highlight the limitations of predictive modelling on a small and imbalanced dataset.p.4