Signal Processing and Classification of Visual Evoked Potentials in a Brain Computer Interface
This research thesis explores an efficient communication approach in P300 based single trial brain-computer interface (BCI). As a tool of rehabilitation engineering for the locked-in patients, the BCI is expected to be swift in performance and intelligent in recognition. With this aim, the objective...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 出版: |
2009
|
| 主題: | |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | This research thesis explores an efficient communication approach in P300 based single trial brain-computer interface (BCI). As a tool of rehabilitation engineering for the locked-in patients, the BCI is expected to be swift in performance and intelligent in recognition. With this aim, the objectives of this thesis are to reduce noise from raw EEG signals using novel variants of principal component analysis (PCA) and singular value decomposition (SVD), to improve classification performance using Fuzzy ARTMAP, Simplified Fuzzy ARTMAP and a combination of other linear classifier and to reduce feature dimension and hardware requirement using genetic algorithm (GA). |
|---|