| Paper: | PS-1A.50 | ||
| Session: | Poster Session 1A | ||
| Location: | H Lichthof | ||
| Session Time: | Saturday, September 14, 16:30 - 19:30 | ||
| Presentation Time: | Saturday, September 14, 16:30 - 19:30 | ||
| Presentation: | Poster | ||
| Publication: | 2019 Conference on Cognitive Computational Neuroscience, 13-16 September 2019, Berlin, Germany | ||
| Paper Title: | Fitting a Computational Model of Perceptual Inference to Principal Component Weights of ERP Responses | ||
| Manuscript: | Click here to view manuscript | ||
| License: |  This work is licensed under a Creative Commons Attribution 3.0 Unported License. |
||
| DOI: | https://doi.org/10.32470/CCN.2019.1214-0 | ||
| Authors: | Lukas Vogelsang, Lilian Weber, Sara Tomiello, Dario Schöbi, Katharina V. Wellstein, Sandra Iglesias, Klaas Enno Stephan, University of Zurich / ETH Zurich, Switzerland | ||
| Abstract: | The mismatch negativity (MMN), a well-studied electrophysiological response to irregularities in the sensory input stream, has often been used to examine how the brain learns the statistics of its environment. This response has also been found to be systematically altered in clinical populations such as patients with schizophrenia. These deviations in electrophysiology, however, cannot easily be linked to inter-individual differences in cognitive processing style due to the lack of direct behavioral readouts, which limits the paradigm’s usefulness for cognitive science and computational psychiatry. To bridge this gap, we present a pipeline for inferring parameters of a generative model of learning and inference, the Hierarchical Gaussian Filter (HGF), given EEG recordings obtained as part of the auditory MMN paradigm. Our pipeline includes a data-driven feature selection step as well as a proposal for mapping belief updates to the EEG features. | ||