h1de0us / mer-eeg-analysis Goto Github PK

This repository is devoted to a research project named "Analysis of the Interaction of Music and Emotions with the Help of EEG"

Before installing the dependencies, create a virtual environment using pyenv or venv and then run

pip install -r requirements.txt

This work uses DEAP dataset, to estimate the connectivity, run

python3 connectivity/src/parse_deap.py --method <desired_method> --data_path <path_to_deap_dataset>

TODO: how to train a model, how to extract features

• connectivity folder describes the process of acquiring graphs from EEG data using various types of connectivity: structural, functional, effective
• graphs folder contains an implementation of a Graphormer model. This structure of the folder is based on a heavily modified fork of this repo, code is partially taken from the official Graphormer implementation
• tda folder contains the code related to the extraction of the topological features of the graph