Using particle swarm to select frequency band and time interval for feature extraction of EEG based BCI

Basis for the implementation of an EEG-based single-trial binary brain computer interface through the disgust produced by remembering unpleasant odors

“…Moreover, efforts will be dedicated to use optimization algorithms, such as the common spatial pattern and its variants [39,40], as classification strategies (on the same data), and to compare them with the proposed method both in accuracy and in efficiency. Time will be dedicated to study other welldistinguishable emotional states to work alongside the disgust to increase the number of choices.…”

“…We used STFT to maintain, of the whole time interval, the most similar periods and to perform the average on these similar data. Data similarity between pieces of signals was evaluated by calculating the r 2 , defined as follows [39,40]:…”

Subject transfer BCI based on Composite Local Temporal Correlation Common Spatial Pattern

“…where the Laplacian matrix L X ¼ D X À W X is a semi-positive definite matrix which can be decomposed as [9]:…”

“…One of the most popular and efficient techniques to extract ERD/ERS related features is the Common Spatial Pattern (CSP), which is widely used for motor imagery BCI designs [6,7]. The CSP method aims to find spatial projections (filters) that simultaneously maximize the variance of one class while minimizing the variance of the other class [8,9]. Despite the efficiency and popularity of CSP in designing BCIs, this algorithm has two inherent drawbacks, one is the high sensitivity to potential outliers and artifacts and the another is the overfitting with small training sets [10].…”

Dimensionality reduction based on distance preservation to local mean for symmetric positive definite matrices and its application in brain–computer interfaces

“…In MDM, first the Riemannian mean of each class is calculated, then a test sample is assigned to the class that has the shortest distance to its mean. Since MDM is not robust to noise, it is suggested to perform some filtering over the data before applying MDM [26]. FGMDM first tries to find a set of filters by applying an extension of Fisher Linear Discriminant Analysis (FLDA) named Fisher Geodesic Discriminant Analysis (FGDA) [12] and then apply these filters to data using a geodesic filtering approach that result in a set of SPD matrices with the same dimensionality as initial data.…”