Built a classification model of the brain during working memory maintenance based on the 128-channel EEG dataset for the material-specific Sternberg task

Extracted features of EEG signals by empirical mode decomposition and phase space reconstruction

Implemented Support Vector Machine (SVM), K-nearest neighbor (KNN), Random Forest (RF), and nested cross-validation to construct the classification model

Improved classification performance from 71.45% to 91.82% under SVM