I wanted to share something I’m working on with the community! I was inspired by the research paper put out last year by Stanford University researchers on NOIR: Neural Signal Operated Intelligent Robots for Everyday Activities. BCI-EEG research is the main reason I started to pursue education in machine learning, and their use of the common spatial pattern algorithm along with quadratic discriminate analysis gave me an idea to test for a multi-label problem. I started a notebook based on an old Kaggle competition to try CSP with different decision tree ensembles to see how they perform with this approach. My notebook is available on Github as a work in progress: Comparative Analysis of Decision Tree Ensemble Models for EEG Event Detection
I appreciate anyone taking a look as I am still new to machine learning engineering, and following best practices. I welcome all suggestions and critiques, especially on how to affordably cut down on training time so I can see some results! 