This work is an implementation of filter-bank common spatial pattern (FBCSP)[1] on BCI Competition IV dataset 2a. While the dataset consist of four class, this work will only use two class which are left and right hand

Early result on train and test data, SVM model is used as classifier, 5-fold CV is used to evaluate model performance

Note : Test result looks suspiciously high, if you guys found something incorrect please open an issue.
I only did for three subjects, but you can extend the result to all nine subjects, the data are there.

[1] Kai Keng Ang, Zheng Yang Chin, Haihong Zhang and Cuntai Guan, "Filter Bank Common Spatial Pattern (FBCSP) in Brain-Computer Interface," 2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence), Hong Kong, 2008, pp. 2390-2397, doi: 10.1109/IJCNN.2008.4634130.
[2] Ang, K. K., Chin, Z. Y., Wang, C., Guan, C., & Zhang, H. (2012). Filter Bank Common Spatial Pattern Algorithm on BCI Competition IV Datasets 2a and 2b. Frontiers in Neuroscience, 6. doi: 10.3389/fnins.2012.00039

2020.11.02 - Train/test result without hp tuning
2024.01.04 - Just general maintenance, fix some stuff, add requirements, dockerfile, devcontainer

I included the datasets, which was originally here https://github.com/bregydoc/bcidatasetIV2a
The datasets is IV2a, the description is here https://www.bbci.de/competition/iv/desc_2a.pdf
This is link to BCI Competition IV https://www.bbci.de/competition/iv/ which contains above page

Future updates (probably...):

• Evaluate on all 9 subjects
• Tune hyperparameters