How can we analyze customer data, when customer data analysis is not allowed any longer?

This and other questions were answered during the second session of the Research Seminar Series with Professor Jaap E. Wieringa, Full Professor of Research Methods in Business at University of Groningen.

Professor Wieringa’s current research presented at WU focuses on the issues of customer data privacy in the light of ever-increasing data exploration for commercial purposes.

The joint work of Prof. Wieringa with his PhD student Gillian Pontepresents ways of using of advanced deep learning methods for generation of synthetic data that mimic real datasets. These methods aim to help companies and institutions avoid potential problems arising from data privacy legislative measures that are becoming more prominent in the wake of recent scandals with customer information leaks. Their presentation showed how General Adversarial Networks (GANs) can be used to successfully create meaningful artificial data with a relatively small training sample, although with certain limitations.

Continuing with the topic of data minimization, Prof. Wieringa presented his and his coauthors’ recent paper on estimation of customer churn with limited present-day data by using a mixture of Kalman filters model, during the doctoral seminar.

We thank Jaap and Gillian for their time and shared knowledge and look forward to another visit!

Alexander Kulumbeg