Abstract
Machine learning methods provide powerful tools to map physical measurements to scientific categories. But are such methods suitable for discovering the ground truth about psychological categories? We use the science of emotion as a test case to explore this question. In studies of emotion, researchers use supervised classifiers, guided by emotion labels, to attempt to discover biomarkers in the brain or body for the corresponding emotion categories. This practice relies on the assumption that the labels refer to objective categories that can be discovered. Here, we critically examine this approach across three distinct datasets collected during emotional episodes—measuring the human brain, body, and subjective experience—and compare supervised classification solutions with those from unsupervised clustering in which no labels are assigned to the data. We conclude with a set of recommendations to guide researchers towards meaningful, data-driven discoveries in the science of emotion and beyond.
Highlights
Machine learning methods provide powerful tools to map physical measurements to scientific categories
Many empirical efforts within psychological science rely on the assumption that Western folk category labels constitute the biological and psychological ‘ground truth’ of the human mind across cultures, while other efforts are more neutral in their assumptions, inferring only that the category characterizes a participant’s behavior
It is possible that the emotion category labels used in these three datasets veridically reflect biological and psychological categories that exist in nature and are stable across contexts and individuals, and that latent emotion constructs do produce distinct, if graded, responses in brain, body, and behavior
Summary
Machine learning methods provide powerful tools to map physical measurements to scientific categories. We re-examine one aspect of the hypothesis that a natural description of the human mind—i.e., the biological organization of behavior and human experience—may require stepping back from the assumption that folk categories describe the ground truth We do this in the context of machine learning to examine how unsupervised and supervised machine learning approaches can be used to build interpretative models from psychology measurements, across three quite different experimental settings. The domain of emotion is a clear test case in which labels that enforce strict category boundaries are traditionally imposed on the data in supervised classification methods Whether these boundaries exist and can be discovered using unsupervised approaches is highly debated (for a discussion, see[19,20]). To the extent that these studies have identified markers that are truly representative of different emotion categories, and generalize across populations and contexts, these studies may be taken by researchers as evidence that emotion categories have a reliable biological and mental basis
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
