Multivariate pattern analyses of Electroencephalography data - Theory and Practice.
23-27th of November 2020
[MULTIVARIATE EEG ANALYSIS] _online course
The course is offered for free to all interested students. The contents, which include theoretical videos, data and Matlab code to implement the analyses, are available on the website for anyone aiming to learn multivariate methods applied to EEG data.
Introduction to the electroencephalography (EEG) signal
- Biological basis and properties.
- EEG oscillations.
- Signal preprocessing for multivariate analyses.
- Considerations for experimental designs.
- Pros and cons of Univariate vs. Multivariate approaches.
Introduction to classifiers
- Families and properties.
- Time-resolved decoding.
- EEG features for classifier training.
- Practice with EEG preprocessing and classification (LDA).
- K-fold cross-validation.
- Searchlight analysis.
- Practice with time generalization.
- Evaluation of the performance of classifiers.
- Controlling methodological and analytical confounds.
- Practice with classifier interpretation and addressing confounds.
Introduction to Representational Similarity Analysis (RSA)
- Practice with RSA in a simple dataset.
- State segmentation methods.
- Beyond multivariate analyses: An introduction to forward encoding models.
- RSA and forward encoding models.
- Practice with building a simple forward encoding model.
Introduction to Deep Neural Networks (DNN)
- Practice with simple DNN.
- Extracting features.
Multivariate analyses in the course will be performed employing Matlab code, so previous familiarity with this programming environment is required. Students without this experience are strongly encouraged to take general introduction courses to Matlab. Enrolled students will need to have Matlab installed in their own computers to be able to follow the theoretical lessons and practical sessions.
This course teaches computer programming to those with little to no previous experience. It uses the programming system and language called MATLAB to do so because it is easy to learn, versatile and very useful for engineers and other professionals.
Join the millions of engineers and scientists who use MATLAB, Simulink, and other add-on products to solve complex design challenges. Are you a student? Your school may provide MATLAB without the 30-day limitation of a trial.
- Under the Mind’s Hood: What We Have Learned by Watching the Brain at Work. J Neurosci. 2020 Jan 2;40(1):89-100. doi: 10.1523/JNEUROSCI.0742-19.2019.
- Haxby JV, Connolly AC, Guntupalli JS. Decoding neural representational spaces using multivariate pattern analysis. Annu Rev Neurosci. 2014;37:435-56. doi: 10.1146/annurev-neuro-062012-170325.
- Kriegeskorte N, Kievit RA. Representational geometry: integrating cognition, computation, and the brain. Trends Cogn Sci. 2013 Aug;17(8):401-12. doi: 10.1016/j.tics.2013.06.007.
- T Deep Neural Networks as Scientific Models. Trends Cogn Sci. 2019 Apr;23(4):305-317. doi: 10.1016/j.tics.2019.01.009.