The introduction of big data analytics in the financial sector has provided bankers with countless opportunities to improve the quality of their services, enhance consumer segmentation, and modernize e-banking.
In my experience in the banking sector, I have utilized data mining techniques to improve the quality of customer segmentation. This approach refers to the extraction of useful data about consumers and its consequent application to solve existing problems (Provost and Fawcett, 2013; Hassani, Huang, and Silva, 2018). Each client has a set of categorical attributes, including demographic, psychographic, financial, behavioral, and transnational factors, that contribute to their consumer conduct (Hassan and Mirza, 2018). In my work, I employed the primary concepts of information gain and entropy to improve customer segmentation among consumers with a negative value – clients whose liabilities exceed their profitability for the bank. I inspected the information entropy values of economic factors – liabilities and creditworthiness – to understand the potential risks of customer interaction. Based on the entropy measures, I modified the values of categorical attributes by defining the information gain and entropy of the category. This approach allowed us to specify the numerical values of categorical attributes (e.g., creditworthiness: 400< FICO/FIRC Score <600), specifically, among customers with a negative value, making the consumer segmentation strategy more accurate.
Consequently, I would like to improve my knowledge concerning data mining techniques and machine learning in banking. For instance, the implementation of innovative applications in entropy analysis, such as EZ Entropy, might significantly improve the speed of customer segmentation analysis (Li, 2019). Consequently, the FIPA method with a focus on information entropy might increase the accuracy of entropy evaluation in categories (Atalay, Atalay, and Isin, 2019). Categorical clustering analysis with the use of entropy parameters is another prominent method to calculate the entropy within segments (Duan, Yang, and Li, 2017). Ultimately, the use of information gain and entropy parameters is an effective strategy for optimizing consumer segmentation, and I would like to improve my knowledge in this area.
Reference List
Atalay, K. D., Atalay, B. and Isin, F. B. (2019). FIPIA with information entropy: A new hybrid method to assess airline service quality. Journal of Air Transport Management, 76, pp. 67-77. Web.
Duan, Q., Yang, Y. L. and Li, Y. (2017). Rough k-models clustering algorithm based on entropy. International Journal of Computer Science, 44(1).
Hassan, M. M. and Mirza, T. (2018). Customer profiling and segmentation in retail banks using data mining techniques. International Journal of Advanced Research in Computer Science, 9(4), pp. 24-29. Web.
Hassani, H., Huang, X. and Silva, E. (2018). Digitalisation and big data mining in banking. Big Data and Cognitive Computing, 2(18).
Li, P. (2019). ‘EZ entropy: A software application for the entropy analysis of physiological time-series. BioMed Online, 18(30).
Provost, F. and Fawcett, T. (2013). Data science for business: What you need to know about data mining and data-analytical thinking. California: O’Reilly.