Publications by authors named "Ronglei Lu"
Article Synopsis
- The study focuses on improving brain-computer interfaces (BCI) for motor imagery (MI) recognition using multi-channel EEG data, highlighting issues like individual differences and noise that can affect performance.
- It introduces a novel channel selection method called Hybrid-Recursive Feature Elimination (H-RFE) that utilizes a combination of recursive feature elimination and various classification techniques (random forest, gradient boosting, logistic regression) to tailor channel selection to individual users.
- Experimental results show significant accuracy improvements in MI recognition, achieving 90.03% on the SHU dataset and 93.99% on the PhysioNet dataset, outperforming traditional channel selection methods by substantial margins.
