Deep Learning: Artificial Neural Network Retail Stores VIP Customers Prediction Study

Abstract

Background/Objective : As recent studies have begun to shed a light on an artificial neural network deep learning based on machine learning techniques, attempts to apply this technology have been increasing in various industries such as production, manufacturing, quality control, marketing, social networks and so on.
Methods/Statistical : In this study, this study analyzes sales data generated from retail stores in Gyeonggi-do, Korea (4 stores), and customer data (approximately 13,000 people) using the stores. In this study, the neural network algorithm was used to show better performance than the existing machine learning techniques. It also shows how to utilize retailer data in multiple data mining techniques using unsupervised learning.
Statistical analysis/Findings : As a result of the analysis, cluster analysis was divided into four clusters, and the model showed the best performance of the neural network model. Of the four clusters, a cluster was created that represented the characteristics of VIP customers and that cluster was modeled as a target. The group's customers were found to be mainly using the store's mobile alarm service and delivery service.
Improvements/Applications : The analysis makes it easy to identify your differentiated marketing and outlook.