OPTIMIZATION OF COSTS FOR MARKETING ACTIVITIES OF BANKING INSTITUTIONS WITH THE HELP OF MODELS BASED ON ENSEMBLES
DOI:
https://doi.org/10.32782/mathematical-modelling/2024-7-1-8Keywords:
marketing activity, banking institutions, model ensembles, cost optimizationAbstract
Marketing activity of banking institutions in the modern world is a key component strategy of their development and competitiveness. Attracting new customers, maintaining existing ones and introducing innovative financial products becomes impossible without an effective marketing strategy. In the conditions of growing competition in the market of banking services, optimization of marketing costs becomes a strategic task for banks The purpose of the study is to use theoretical knowledge to develop effective models based on ensembles using parameter tuning to optimize the costs of marketing activities of banking institutions, which will reduce marketing costs and increase the effectiveness of marketing activities. The article examines an approach to optimizing the costs of marketing activities of banking institutions using several types of ensemble models. In particular, random forest, gradient boosting, extreme gradient boosting. The authors investigate the marketing activities of banks and how to effectively use ensemble-based models in RStudio environment to achieve maximum results, using a dataset with data from a Portuguese bank's marketing campaign. Detailed analysis includes examining factors collected by the bank about customers and evaluating their impact on the final analysis of the customer's decision. Also, in the research process, the methods of under-sampling, scaling and finding the optimal number of features are used to improve the modeling results. As a result of the study, it was found that the effectiveness of the marketing campaign when using models is actually 2 times higher than without them. The results of the study are a valuable guide for banking institutions to optimize their marketing activities. The use of model ensembles can contribute to increasing the effectiveness of marketing senses, reducing costs and increasing competitiveness in the financial services market.
References
Safarkhani F., Moro S. Improving the Accuracy of Predicting Bank Depositor’s Behavior Using a Decision Tree. Applied Sciences. 2021. 11(19). 9016.
Hao Ruiting, Xia Xiaoqian, Shen Siyi, Yang Xiaorong. Bank Direct Marketing Analysis Based on Ensemble Learning. Journal of Physics: Conference Series, 2020.
Chen X.-W., Jeong J.C. Enhanced recursive feature elimination, Sixth International Conference on Machine Learning and Applications (ICMLA 2007), Cincinnati, OH, USA, 2007, pp. 429–435, doi: 10.1109/ICMLA.2007.35.
Jović A., Brkić K., Bogunović N. A review of feature selection methods with applications, 2015 38th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO), Opatija, Croatia, 2015, pp. 1200–1205.
Meidan A. Bank Marketing Strategies, International Journal of Bank Marketing, 1983. Vol. 1. No. 2, pp. 3–17.
González-Recio O., Jiménez-Montero J.A., Alenda R. The gradient boosting algorithm and random boosting for genome-assisted evaluation in large data sets, Journal of Dairy Science, 2013. Volume 96, Issue 1, p. 614–624.
Tianqi Chen and Carlos Guestrin. 2016. XGBoost: A Scalable Tree Boosting System. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD '16). Association for Computing Machinery, New York, NY, USA, p. 785–794.
Bunkhumpornpat C., Sinapiromsaran K., Lursinsap C. MUTE: Majority under-sampling technique, 2011 8th International Conference on Information, Communications & Signal Processing, Singapore, 2011, pp. 1–4.
Moro S., Rita P., Cortez P. Bank Marketing. UCI Machine Learning Repository. 2012.
Alibrahim Hussainand, Ludwig Simone A. Hyperparameter Optimization: Comparing Genetic Algorithm against Grid Search and Bayesian Optimization. In 2021 IEEE Congress on Evolutionary Computation (CEC). IEEE Press, 2021, p. 1551–1559.
Amalia Luque, Alejandro Carrasco, Alejandro Martín, Ana de las Heras, The impact of class imbalance in classification performance metrics based on the binary confusion matrix, Pattern Recognition, Volume 91, 2019, Pages 216–231.
