Predicting customer choice: A case study on integrating AI within a discrete choice model | Kathryn
About This Video
Neural networks have been widely celebrated for their power to solve difficult problems across a number of domains. We explore an approach for leveraging this technology within a statistical model of customer choice. Conjoint-based choice models are used to support many high-value decisions at GM. In particular, we test whether using a neural network to model customer utility enables us to better capture non-compensatory behavior (i.e., decision rules where customers only consider products that meet acceptable criteria) in the context of conjoint tasks. We find the neural network can improve hold-out conjoint prediction accuracy for synthetic respondents exhibiting non-compensatory behavior only when trained on very large conjoint data sets. Given the limited amount of training data (conjoint responses) available in practice, a mixed logit choice model with a traditional linear utility function outperforms the choice model with the embedded neural network.
This workshop was conducted by Kathryn Schumacher, Staff Researcher in the Advanced Analytics Center of Expertise within General Motor’s Chief Data and Analytics Office.
In This Video
Staff Researcher, Chief Data and Analytics Office, General Motors
Kathryn Schumacher is Staff Researcher in the Advanced Analytics Center of Expertise within General Motor’s Chief Data and Analytics Office. Kathryn earned her MS and PhD in Industrial and Operations Engineering at the University of Michigan in 2011 and 2014, respectively. She has a BS in Chemical Engineering from MIT.
Kathryn’s work at GM focuses on using discrete choice models to improve decision making. She has led projects on optimizing vehicle prices and incentives, and on using AI for choice modeling. She has five inventions recognized as GM Trade Secrets. She is a two-time winner of GM’s highest internal award for innovations with demonstrated business impact.