Marketing Database augmentation and fusion using deep learning

Abstract

The authors have proposed and being conducting a method for adapting deep learning-based data augmentation and fusion to human-factors / ergonomic data. In this study, we have experimentally examined the adaptation of our proposed method to questionnaire and behavioral data used for marketing purposes. The results showed that augmentation and fusion of two different marketing databases (questionnaire and behavioral log data) using GAN had effects of improving the prediction accuracy of specific target variables (e.g. purchased products).