Best-Classifier Feedback In Diagnostic Classification Training
Keywords
Classification; Diagnosis; Feedback; Optimal-classifier; Training
Abstract
Diagnostic classification training requires viewing many examples along with category membership feedback. "Objective" feedback based on category membership suggests that perfect accuracy is attainable when it may not be (e.g., with confusable categories). Previous work shows that feedback based on an "optimal" responder (that sometimes makes classification errors) leads to higher long-run reward, especially in unequal category payoff conditions. In the current study, participants learned to classify normal or cancerous mammography images, earning more points for correct "cancer" than "normal" responses. Feedback was either objective or based on performance of an empirically determined "best" classifier. This approach is necessary because theoretically optimal responses cannot be determined with complex real-world stimuli with unknown perceptual distributions. Replicating earlier work that used simple artificial stimuli, we found that best-classifier performance led to decision-criterion values (β) closer to the reward-maximizing criterion, along with higher point totals and a slight reduction (as predicted) in overall accuracy.
Publication Date
12-1-2015
Publication Title
Journal of Applied Research in Memory and Cognition
Volume
4
Issue
4
Number of Pages
368-373
Document Type
Article
Personal Identifier
scopus
DOI Link
https://doi.org/10.1016/j.jarmac.2015.07.007
Copyright Status
Unknown
Socpus ID
84939832835 (Scopus)
Source API URL
https://api.elsevier.com/content/abstract/scopus_id/84939832835
STARS Citation
Bohil, Corey J.; Wismer, Andrew J.; Schiebel, Troy A.; and Williams, Sarah E., "Best-Classifier Feedback In Diagnostic Classification Training" (2015). Scopus Export 2015-2019. 779.
https://stars.library.ucf.edu/scopus2015/779