Sign inSign up
Gentle Introduction to Signal Processing and Classification for Single-Trial EEG Analysis thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Gentle Introduction to Signal Processing and Classification for Single-Trial EEG Analysis

Published on Jun 17, 20136975 Views

Benjamin Blankertz

The aim of this lecture is to provide an illustrative tutorial on the methods for single-trial EEG analysis. Basic concepts of feature extraction and classification will be explained in a way that is

BBCI Summer School

Related categories

NeuroscienceNeurotechnology

Chapter list

Gentle Introduction to Signal Processing and Classication for Single-Trial EEG Analysis00:00
Univariate Features: Averages and Single-Trials00:05
Receiver Operator Characteristics (ROC) and AUC (1)01:54
Receiver Operator Characteristics (ROC) and AUC (2)02:53
Receiver Operator Characteristics (ROC) and AUC (3)03:18
Receiver Operator Characteristics (ROC) and AUC (4)03:31
Receiver Operator Characteristics (ROC) and AUC (5)03:40
Receiver Operator Characteristics (ROC) and AUC (6)04:08
Receiver Operator Characteristics (ROC) and AUC (7)04:47
Receiver Operator Characteristics (ROC) and AUC (8)05:08
From Uni- to Multivariate Features05:56
Multi-channel Epochs06:43
The Virtue of Multivariate Spatial Features (1)07:12
The Virtue of Multivariate Spatial Features (2)07:47
The Virtue of Multivariate Spatial Features (3)08:29
The Virtue of Multivariate Spatial Features (4)08:50
The Virtue of Multivariate Spatial Features (5)09:30
The Virtue of Multivariate Spatial Features (6)09:53
The Virtue of Multivariate Spatial Features (7)10:25
The Virtue of Multivariate Spatial Features (8)10:45
ERPs in a Head Plot11:43
Interlude: Representation as Matrix (1)12:49
Interlude: Representation as Matrix (2)13:22
ERPs in a Grid Plot15:32
ERP Topographies (1)16:33
ERP Topographies (2)17:56
AUC Matrix: Selection of Channels and Time Intervals19:02
Multivariate ERP Features (1)20:23
Multivariate ERP Features (2)20:46
Multivariate ERP Features (3)20:51
Representation of Multivariate Distributions: Scatter Plot (1)21:58
Representation of Multivariate Distributions: Scatter Plot (2)22:51
Representation of Multivariate Distributions: Scatter Plot (3)23:11
Representation of Multivariate Distributions: Scatter Plot (4)23:35
Representation of Multivariate Distributions: Scatter Plot (5)23:57
Representation of Multivariate Distributions (2)24:22
Distribution of the Noise24:43
Distribution of ERP Features25:01
Two Univariate Gaussian Distributions28:33
Two-Dimensional Gaussians - Correlated or Uncorrelated29:07
Correlated or Uncorrelated? Mind Spatial Smearing!30:12
Gaussian Distributions31:02
Eigenvalue Decomposition32:11
Characterization of Gaussian Distributions33:20
Nearest Centroid Classier (NCC) -135:35
Nearest Centroid Classier (NCC) -236:15
Nearest Centroid Classier (NCC) -336:22
Nearest Centroid Classier (NCC) -436:46
Linear Disciminant Analysis (1)39:15
Linear Disciminant Analysis (2)39:43
Linear Disciminant Analysis (3)41:23
Linear Disciminant Analysis (4)41:57
Mean and Eigenvalue Spectrum for a P300 Data Set45:31
The Structure of the Noise45:37
For Comparison: Covariances in Handwritten Digits46:37
Validation of Classication Procedures48:20
Loss Function for Unbalanced Classes49:56
Application of (Purely) Temporal Features52:10
Application of (Purely) Spatial Features (1)53:25
Application of (Purely) Spatial Features (2)54:17
Classication of Spatio-Temporal Features (1)55:10
Classication of Spatio-Temporal Features (2)55:11
Results of Classifying Spatial Features55:25
Bias in Estimating Covariance Matrices (1)55:39
Bias in Estimating Covariance Matrices (2)56:15
Bias in Estimating Covariances (2)57:00
A Remedy for the Estimation Bias58:18
Properties of the Shrunk Covariance Matrix (1)58:59
Properties of the Shrunk Covariance Matrix (2)59:03
Regularized Linear Discriminant Analysis59:34
Impact of Shrinkage as Trade-off (1)01:00:47
Impact of Shrinkage as Trade-off (2)01:00:49
Regularized LDA at Work01:01:08
Optimal Selection of Shrinkage Parameter (1)01:01:35
Optimal Selection of Shrinkage Parameter (2)01:02:00
Classication with Shrinkage-LDA01:03:12
Classication on Single Components and Combined (1)01:04:14
Classication on Single Components and Combined (2)01:04:39
Recap: Classication of (Purely) Spatial Features01:06:24
Interpretation of Spatial Filters01:11:46
Interpretation of Spatial Filters (2)01:12:02
Understanding Spatial Filters (1)01:12:42
Understanding Spatial Filters (2)01:13:14
References01:14:23