Development of Machine Learning based Cash Forecasting Models for Automated Teller Machines

Abstract

Nowadays, Automatic Teller Machines (ATMs) stand out as bank instruments where a high percentage of cash transactions take place. Accurately determining the amount of cash to be kept in ATMs is considered a strategic necessity for banks in terms of preventing service disruptions and maximizing customer satisfaction. Cash forecasting ensures that the amount of cash to be kept in ATMs is determined accurately. The aim of this study is to optimize cash management by forecasting daily cash demand in ATMs and thus help financial institutions prevent inefficiencies caused by cash depletion in ATMs and reduce customer dissatisfaction and operational costs. To achieve this, cash forecasting models have been developed using Extreme Gradient Boosting (XGBoost). The performance of the models has been evaluated with the Percentage Error (PE) metric. The developed models provided error values lower than 15%. A comprehensive evaluation has shown that accurate cash forecasts significantly increase the effectiveness of cash management.

Keywords

• Cash Forecasting
• Cash Management
• ATM
• XGBoost

References

1. [1] Mubiru, K. P., Ssempijja, M. N. (2024). Modelling and optimization of ATM cash-loading under stochastic demand. Discover Analytics; 2: 14. https://doi.org/10.1007/s44257-024-00023-0
2. [2] Sarveswararao, V., Ravi, V., Vivek, Y. (2023). ATM cash demand forecasting in an Indian bank with chaos and hybrid deep learning networks. Expert Systems with Applications; 211: 118645. https://doi.org/10.1016/j.eswa.2022.118645
3. [3] Bartzsch, N., Brandi, M., Devigne, L., De Pastor, R., Maddaloni, G., Posada Restrepo, D., Sene, G. (2023). Forecasting Euro Banknotes in Circulation with Structural Time Series Models in Times of the COVID-19 Pandemic.
4. [4] Orhan, D., Erol Genevois, M. (2023, August). Cash Replenishment and Vehicle Routing Improvement for Automated Teller Machines. In International Conference on Intelligent and Fuzzy Systems, Cham: Springer Nature Switzerland; 758: 721-729. https://doi.org/10.1007/978-3-031-39774-5_80
5. [5] Thanh, B. T., Van Tuan, D., Chi, T. A., Van Dai, N., Dinh, N. T. Q., Thuy, N. T., Hoa, N. T. X. (2023). Multiobjective Logistics Optimization for Automated ATM Cash Replenishment Process. In International Conference on Intelligence of Things, Cham: Springer Nature Switzerland; 187: 46-56. https://doi.org/10.1007/978-3-031-46573-4_5
6. [6] Fallahtafti, A., Aghaaminiha, M., Akbarghanadian, S., Weckman, G. R. (2022). Forecasting ATM cash demand before and during the COVID-19 pandemic using an extensive evaluation of statistical and machine learning models. SN computer science; 3: 164. https://doi.org/10.1007/s42979-021-01000-0
7. [7] Suder, M., Gurgul, H., Barbosa, B., Machno, A., Lach, Ł. (2024). Effectiveness of ATM withdrawal forecasting methods under different market conditions. Technological Forecasting and Social Change; 200: 123089. https://doi.org/10.1016/j.techfore.2023.123089
8. [8] Cedolin, M., Genevois, M. E. (2024). Cash Demand Prediction Problem using Econometric and Computational Intelligence Forecasting Models. In 2024 10th International Conference on Control, Decision and Information Technologies, IEEE, 103-108. https://doi.org/10.1109/CoDIT62066.2024.10708473

9. [9] Cedolin, M., Orhan, D., Genevois, M. (2024). Statistical and Artificial Intelligence Based Forecasting Approaches for Cash Demand Problem of Automated Teller Machines. Academic Platform Journal of Engineering and Smart Systems; 12: 21-27. https://doi.org/10.21541/apjess.1360151
10. [10] Gurgul, H., Lach, Ł., Suder, M., Szpyt, K. (2023). Using trigonometric seasonal models in forecasting the size of withdrawals from automated teller machines. Entrepreneurial Business and Economics Review; 11: 181-204. http://dx.doi.org/10.15678/EBER.2023.110312
11. [11] Kumar, H., Megavarshini, G., Sreenivasan, A. (2023). Machine Learning in the Banking Sector, Proceedings of the International Conference on Industrial Engineering and Operations Management Manila, 1081-1088. https://doi.org/10.46254/AN13.20230313
12. [12] Zeinalkhani, M., Ghanbar Tehrani, N., Pasandideh, S. H. R., Pedram, M. M. (2024). Providing a forecasting model and optimization of the cash balance of bank branches and ATMs with the approach of social responsibilities. International Journal of Nonlinear Analysis and Applications; 15: 211-224.
13. [13] Nigmatulin, G. A., Chaganova, O. B. (2022). Research of an optimization model for servicing a network of ATMs and information payment terminals. arXiv, 2210.09927. https://doi.org/10.48550/arXiv.2210.09927
14. [14] Özlem, Ş., Tan, O. F. (2022). Predicting cash holdings using supervised machine learning algorithms. Financial Innovation; 8: 44. https://doi.org/10.1186/s40854-022-00351-8
15. [15] Balakrishnan, T. V., Kalaiarasi, R. (2021). A Literature Survey On Automated Teller Machine Cash Demand Analysis And Prediction In Financial Sector. Ilkogretim Online; 20: 2146-2161.
16. [16] Cedolin, M., Erol Genevois, M. (2022). An averaging approach to individual time series employing econometric models: a case study on NN5 ATM transactions data. Kybernetes; 51: 2673-2694. https://doi.org/10.1108/K-03-2021-0235
17. [17] Gorodetskaya, O., Gobareva, Y., Koroteev, M. (2021). A machine learning pipeline for forecasting time series in the banking sector. Economies; 9: 205. https://doi.org/10.3390/economies9040205
18. [18] Peykani, P., Eshghi, F., Jandaghian, A., Farrokhi-Asl, H., Tondnevis, F. (2021). Estimating cash in bank branches by time series and neural network approaches. Big data and computing visions; 1: 170-178. https://doi.org/10.22105/BDCV.2021.142232
19. [19] Rafi, M., Wahab, M. T., Khan, M. B., Raza, H. (2021). Towards optimal ATM cash replenishment using time series analysis. Journal of Intelligent and Fuzzy Systems; 41: 5915-5927. https://doi.org/10.3233/JIFS-201953 [20] Chavan, V. V., Manjaly, J. J., Ali, M. A. (2021). Automated Teller Machine Cash Demand Prediction. In 2021 IEEE 6th International Forum on Research and Technology for Society and Industry, IEEE, 142-147. https://doi.org/10.1109/RTSI50628.2021.9597352
20. [21] Park, S., amp;amp; Kim, J. (2019). Landslide susceptibility mapping based on random forest and boosted regression tree models, and a comparison of their performance. Applied Sciences; 9: 942. https://doi.org/10.3390/app9050942
21. [22] Rad, B. B., Bhatti, H. J., Ahmadi, M. (2017). An introduction to docker and analysis of its performance. International Journal of Computer Science and Network Security (IJCSNS); 7: 228.
22. [23] Shamim, S. I., Gibson, J. A., Morrison, P., Rahman, A. (2022). Benefits, Challenges, and Research Topics: A Multi-vocal Literature Review of Kubernetes. arXiv, 2211.07032. https://doi.org/10.48550/arXiv.2211.07032