Challenges for Analysis and Visualization using encoding models of EEG brain activity
Event date:  December 7, 2020, 4:00 PM  to 6:00 PM


Human brain potentials have been shown to be helpful for Human Computer Interactions (HCI) both as a direct input channel (Brain-CI), as well as a tool to understand user behavior. But the analysis of such data can be quite challenging as soon as classical experimental restrictions of interaction are removed. For many natural human-computer interactions, eye- and hand-movements are elemental.

The analyses of such data are challenging because we need statistical control for many continuous confounders, and we need deconvolution methods due to temporal overlap of brain potentials between gaze-fixations. To overcome these challenges, we recently introduced an integrated workflow based on encoding models and time-regression.

The visualizations of such encoding models are complex, mainly because the encoding model results reside in high-dimensional space (sensors x timepoints x subjects x predictors). This makes selection or discretization of the space necessary. In addition, the brain activity is non-stationary over time, complicating averaging over timepoints and subjects show high variability to each other, complicating averaging on the group level. Further, the experimental predictors are often not orthogonal but can potentially interact in surprising ways (e.g. suppressor effects, mediation).

In this talk I will give an introduction to the topic, intuit the logic of the analyses and discuss some problems in visualization.

About the Speaker

2018 - PhD in Cognitive Science, Osnabrück, Germany

2018-2020: PostDoc in PredictiveBrainLab with Floris deLange at Donders Institute, Nijmegen, Netherlands

2020: Tenure-Track Professor Computational Cognitive Science


About the Series

The Lecture Series is organized every winter term and consist of talks with international speakers. The speakers are experts from various fields and professions. The talks are aimed at covering the whole spectrum of visual computing and at discussing the relevance of quantification. This years lecture series will be held entirely online.

This is an online event only.

Registration: If you would like to attend, please email Leonel Merino to register.