Ruiz-Blondet, M.*, Khalifian, N., Armstrong, B. C.,
Jin, Zanpeng, J., Kurtz, K. J., Laszlo, S. (2014).
Brainprint: Identifying unique features of neural activity with
machine learning. Proceedings of the 36th Annual Conference of the Cognitive Science Society (pp. 827-832). Mahwah, NH: Lawrence Erlbaum Associates.
Download:
Author's self-archived version (.pdf) (6 pages)
Official version hosted by the Cognitive Science Society [External Link]
Abstract
Can a person be identified uniquely by some feature of their neural activity, as they can be by fingerprints? If so, 1) what would those features be like and 2) are existing computational methods sufficient to extract them? Here, we explore these questions by coordinating psychophysiological and machine learning approaches. We begin with the proposition that one unique feature of individual cognition is the detailed network of concepts, and relationships between concepts, that are present in each individual’s semantic memory. We then demonstrate that we are able to accurately classify individual unlabeled brain activity—in the form of Event-Related Potentials (ERPs) elicited during a task that probes semantic memory—to the individual it belongs to with several pattern classifiers. These results demonstrate that it is possible to identify individuals on the basis of unique features of their brain activity. Biometric applications are discussed.
Keywords: Machine Learning; Event-Related Potentials; Individual Differences; Biometrics
Copyright Notice (borrowed from David Plaut): The documents distributed here have been provided as a means to ensure timely dissemination of scholarly and technical work on a noncommercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.