September 2015
Volume 15, Issue 12
Free
Vision Sciences Society Annual Meeting Abstract  |   September 2015
Classifying EEG patterns of visual statistical learning
Author Affiliations
  • Brett Bays
    Department of Psychology, University of California, Riverside
  • Aaron Seitz
    Department of Psychology, University of California, Riverside
Journal of Vision September 2015, Vol.15, 387. doi:10.1167/15.12.387
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Brett Bays, Aaron Seitz; Classifying EEG patterns of visual statistical learning. Journal of Vision 2015;15(12):387. doi: 10.1167/15.12.387.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Statistical learning (SL) refers to the extraction of probabilistic relationships between stimuli and is increasingly used as a method to understand learning processes. However, little is known regarding how SL accumulates over time, nor of the neural processes that underlie SL. To address these issues, in this study we employ pattern classification techniques to examine electroencephalography (EEG) data collected as participants acquire SL. While recording EEG, we exposed participants to a stream of visual shapes which, unbeknownst to them, were grouped into pairs, and then subsequently tested for statistical learning using a reaction time based search task. We then use a k-Nearest Neighbors pattern classification algorithm to classify corresponding EEG signals under two test conditions: classifying periods of activity after stimuli appear based on the presentation statistics of those stimuli; and classifying periods of activity after stimuli appear based on whether those stimuli correspond to "learned" or "non-learned" behavioral patterns. With these tests we show that by using behavioral measures to label certain items as "learned" and other items as "non-learned", we can design a classifier that is able to successfully discriminate those patterns of EEG activity within participants. In future work, we hope that these classifiers can be adapted to the online analysis of SL so as to detect SL as it is acquired. This has important implications for the field of SL and is a step toward understanding how SL accumulates over time and the neural processes that underlie SL.

Meeting abstract presented at VSS 2015

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×