Related Experiment Video
Updated: May 13, 2026
08:51
Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms
Published on: November 1, 2019
Characterizing multivariate decoding models based on correlated EEG spectral features.
1Laboratory of Neural Injury and Repair, Wadsworth Center, New York State Department of Health, P.O. Box 509, Empire State Plaza, Albany, NY 12201-0509, USA. mcfarlan@wadsworth.org
Summary
Multivariate decoding models improve cursor control prediction but can be difficult to interpret when using correlated predictors. Careful interpretation and comparison with univariate statistics are recommended for these neurophysiological analyses.
Area of Science:
- Neuroscience
- Machine Learning
- Signal Processing
Background:
- Multivariate decoding is widely used for analyzing neurophysiological data.
- Interpreting these models can be challenging, especially with correlated predictors.
Purpose of the Study:
- To investigate interpretative issues in multivariate decoding when predictors are correlated.
- To compare the predictive and interpretive performance of univariate and multivariate models.
Main Methods:
- Analysis of sensorimotor rhythm data using linear univariate and multivariate models.
- Feature extraction via autoregressive (AR) spectral analysis with varying model orders.
- Introduction of multicollinearity through AR feature selection.
Main Results:
- Multivariate models significantly outperformed univariate models in predicting target position.
- Interpretation of spectral patterns in model weights was hindered by high multicollinearity in lower-order AR features.
- The predictive power of multivariate models was high, but interpretability suffered.
Conclusions:
- Exercise caution when interpreting multivariate model weights with correlated predictors.
- Comparing results with univariate statistics is advisable for robust interpretation.
- Multivariate decoding is powerful for prediction but may have limited interpretability with multicollinearity.
Related Concept Videos
01:25Correlation between ECG and Cardiac Cycle
The electrical signals recorded on an electrocardiogram (ECG) occur before the mechanical processes of contraction and relaxation during the cardiac cycle.
A cardiac action potential originates in the SA node and spreads throughout the atria and the AV node in approximately 0.03 seconds. This results in the P wave in an ECG and triggers atrial contraction. The action potential is then briefly slowed at the AV node, allowing the atria to contract and fill the ventricles with blood before...
A cardiac action potential originates in the SA node and spreads throughout the atria and the AV node in approximately 0.03 seconds. This results in the P wave in an ECG and triggers atrial contraction. The action potential is then briefly slowed at the AV node, allowing the atria to contract and fill the ventricles with blood before...
01:30Classification of Signals
In signal processing, signals are classified based on various characteristics: continuous-time versus discrete-time, periodic versus aperiodic, analog versus digital, and causal versus noncausal. Each category highlights distinct properties crucial for understanding and manipulating signals.
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...