Predicting Consumer Choice from Raw Eye-Movement Data using the RETINA Deep Learning Architecture
COM1 Level 2
SR1, COM1-02-06
Abstract:
In this talk we present a deep learning architecture, called RETINA, to predict multi-alternative consumer choice from the eye movement data. Unlike previous research, RETINA uses raw eye-tracking data from both eyes as input. It combines the BERT and the Metric Learning methods which capitalize on the key characteristics of the raw eye-tracking data. Using the raw data input eliminates the information loss that may result from first calculating fixations, deriving metrics from the fixations data and analyzing those metrics, as is usually done in the eye movement research. We tested the RETINA architecture on a data set with 112 users who made choices among four laptops, we show that the proposed architecture outperforms other state-of-the-art machine learning approaches, such as standard BERT, LSTM, AutoML and logistic regression methods, that were calibrated on raw or fixation data. Using AutoML, we provide an assessment of which features of the eye movement data contribute to the prediction of consumer choice behavior. The analysis of partial time and partial data segments reveals the ability of RETINA to predict choice outcomes well before a choice decision has been reached. Finally, we provide recommendations on how the proposed deep learning architecture can be used in practice.
Bio:
Alex Tuzhilin is Leonard N. Stern Professor of Business in the Department of Technology, Operations and Statistics at the Stern School of Business, NYU. His research interests include personalization, recommender systems, machine learning and AI. He has produced over 160 research publications on these and related topics. Professor Tuzhilin has served on the organizing committees of numerous conferences, including as the Program and the General Chair of the IEEE International Conference on Data Mining (ICDM), and as the Program and the Conference Chair of the ACM Conference on Recommender Systems (RecSys). He served on the editorial boards of several journals, including as the Editor-in-Chief of the ACM Transactions on Management Information Systems. Professor Tuzhilin's past doctoral students have joined leading universities, including the Wharton School, University of Minnesota, Georgia Institute of Technology, Boston University, Emory University, Temple University and the Hong Kong University of Science and Technology. As a result of various academic accomplishments, the ISS INFORMS Society has honored him with the title of Distinguished Fellow.