Deep and Wide Transfer Learning with Kernel Matching for Pooling Data from Electroencephalography and Psychological Questionnaires

Abstract

Motor imagery (MI) promotes motor learning and encourages brain–computer interface systems that entail electroencephalogram (EEG) decoding. However, a long period of training is required to master brain rhythms’ self-regulation, resulting in users with MI inefficiency. We introduce a parameter-based approach of cross-subject transfer-learning to improve the performances of poor-performing individuals in MI-based BCI systems, pooling data from labeled EEG measurements and psychological questionnaires via kernel-embedding. To this end, a Deep and Wide neural network for MI classification is implemented to pre-train the network from the source domain. Then, the parameter layers are transferred to initialize the target network within a fine-tuning procedure to recompute the Multilayer Perceptron-based accuracy. To perform data-fusion combining categorical features with the real-valued features, we implement stepwise kernel-matching via Gaussian-embedding. Finally, the paired source–target sets are selected for evaluation purposes according to the inefficiency-based clustering by subjects to consider their influence on BCI motor skills, exploring two choosing strategies of the best-performing subjects (source space): single-subject and multiple-subjects. Validation results achieved for discriminant MI tasks demonstrate that the introduced Deep and Wide neural network presents competitive performance of accuracy even after the inclusion of questionnaire data.

Psychological Analysis of Athletes during Basketball Games from the Perspective of Deep Learning

“…Players do not participate in the competition for a long time and have no experience in the competition. Every time after a major competition, their mentality is easy to be too nervous and thus reduce the scoring and shooting rate [18].…”

Subject-Dependent Artifact Removal for Enhancing Motor Imagery Classifier Performance under Poor Skills

“…On the other hand, Table 2 displays the resulting performance obtained through a 10-fold cross-validation scheme, as commonly validated for DBII [ 53 ]. Unfortunately, as shown in the central column, both approaches (i.e., Ind-AR ) result in the worst bi-class classification scenario and harm average over all individuals rather than improving the performance, as expected.…”

KCS-FCnet: Kernel Cross-Spectral Functional Connectivity Network for EEG-Based Motor Imagery Classification

“…Additionally, further analysis of the inter-class variability will be conducted, specifically focusing on how each subject performs within different runs of the experiment. We aim to understand if subjects learn to execute motor imagery tasks as they progress through the runs and whether this results in better performance in subsequent runs [ 52 , 53 ]. Furthermore, we plan to analyze subject-specific filters, specifically looking at the activity of the filter waves concerning brain frequency bands such as Theta, Alpha, Beta, and Gamma.…”