Abstract:
Due to the low signal-to-noise ratio of scalp EEG, brain–computer interface (BCI) development requires accurate electroencephalogram (EEG) pattern decoding for specific mental tasks. Machine learning can optimize EEG patterns for better decoding. Existing algorithms fail to explore the underlying data structure capturing the true EEG sample distribution, resulting in suboptimal decoding accuracy. A clustering-based multitask feature learning algorithm for EEG pattern decoding can reveal EEG data’s intrinsic distribution structure. We use affinity propagation-based clustering to identify subclasses (clusters) in each of the original classes and then label each subclass using a one-versus-all encoding strategy. We use the encoded label matrix and the subclass relationship to jointly optimize EEG pattern features from the uncovered subclasses in a novel multitask learning algorithm. We train a linear support vector machine with optimized features for EEG pattern decoding. Three EEG data sets are used for extensive experimental studies to compare our algorithm to other state-of-the-art methods. Our algorithm is superior for EEG pattern decoding in BCI applications, as shown by improved experimental results.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
