|
| Signal processing | Feature extraction | Machine learning algorithms |
|
| (i) Channel selection | (i) Multiwindow averages [26, 27] | (i) Linear discriminant Analysis (LDA) [28] |
| (ii) Resampling | (ii) Common Spatial Patterns (CSP) [29] | (ii) Quadratic discriminant analysis (QDA) [30] |
| (iii) Artifact rejection (spike detection, bad window detection, bad channel detection, local peak detection) | (iii) Spectrally-weighted common spatial patterns [31] | (iii) Regularized and analytically regularized LDA and QDA [30, 32] |
| (iv) Envelope extraction | (iv) Adaptive autoregressive modeling, from BioSig [33] | (iv) Linear SVM [34] (LIBLINEAR/CVX) |
| (v) Epoch extraction | (1) Dual-agumented lagrange (DAL) [25] | (v) Kernel SVM [34] |
| (1) Time-frequency window selection | (2) Frequency-domain DAL (FDAL) | (vi) Gaussian mixture models (GMM), 9 methods [35–37]) |
| (2) Spectral transformation | (3) Independent Modulators [38] | (vii) Regularized and variational Bayesian logistic regression and sparse Bayesian logistic regression [39, 40] |
| (vi) Baseline filtering | (4) Multiband-CSP [41] | (1) Hierarchical kernel learning [42] |
| (vii) Resampling | (5) Multi-Model Independent component features | (viii) Relevance vector machines (RVM) [43] |
| (viii) Re-referencing | | (1) group-sparse/rank-sparse linear and logistic regression [25] |
| (ix) Surface Laplacian filtering [44] | | (2) high-dimensional Gaussian Bayes density estimator/classifier |
| (x) ICA methods (Infomax, FastICA, AMICA) [6, 45] | | (3) Voting metalearner |
| (xi) Spectral filters (FIR, IIR) | | |
| (xii) Spherical spline interpolation [46] | | |
| (1) Signal normalization | | |
| (2) Sparse signal reconstruction (NESTA, SBL [47], FOCUSS, l1; currently offline only) | | |
| (3) Linear projection | | |
|
