Genetic Algorithm-Based Adaptive Pattern Mining for Customer Basket Analysis

Abstract

Customer basket analysis aims to understand purchasing patterns and enhance marketing strategies by leveraging data mining techniques. To the best of our knowledge, this study is one of the few that applies low-support adaptive pattern mining in personalization, introducing a robust genetic algorithm for improving static transaction baskets.The proposed method uses GA integrated with sophisticated pattern mining methods that quickly finds complex associations as well as patterns from transaction data. By utilizing GAs, our model has the ability to adapt and evolve to fit this changing landscape of consumer purchasing behaviors. The GA-based method is compared to traditional data mining methods using a large dataset of retail transactions. Further it shows big improvements in both accuracy and speed. The results indicate that the adaptive nature of the proposed method not only helps uncover purchasing patterns but also supports retailers in making informed decisions on inventory control, promotional activities, and personalized marketing. The findings provide a promising insight into how implementing genetic algorithms in combination with pattern mining frameworks could yield more proactive and responsive customer analysis within the industry of retail.