On the need for on-line learning in brain-computer interfaces

Millán, José del R.

doi:10.1109/ijcnn.2004.1381116

Cited by 145 publications

(125 citation statements)

References 19 publications

Supporting

Mentioning

122

Contrasting

Unclassified

Order By: Relevance

“…• Graz dataset: Dataset III in BCI competition II, which was provided by the Laboratory of Brain-Computer Interfaces (BCI-Lab), Graz University of Technology (Blankertz et al, 2004;Lemm et al, 2004); • IDIAP dataset: Dataset V in BCI competition III, which was provided by the IDIAP Research Institute (J. del R. Millán, 2004).…”

Section: Eeg Classificationmentioning

confidence: 99%

Non-negative matrix factorization with α-divergence

Cichocki

Lee

Kim

et al. 2008

Pattern Recognition Letters

121

View full text Add to dashboard Cite

Section: Eeg Classificationmentioning

confidence: 99%

Non-negative matrix factorization with α-divergence

Cichocki

Lee

Kim

et al. 2008

Pattern Recognition Letters

121

View full text Add to dashboard Cite

“…Several scaling factors were used to translate these features into positions on the screen, four of which were successively adapted to the individual user during the session. Similarly, [20] investigated a scenario involving a four-class BCI classification problem. The estimation of means and covariance matrices for each of the classes was iteratively updated in a simulated online scenario; these parameter changes indicated the possibility of considerable improvement for online control.…”

Section: Introductionmentioning

confidence: 99%

Towards adaptive classification for BCI

Shenoy

Krauledat

Blankertz³

et al. 2006

J. Neural Eng.

385

347

View full text Add to dashboard Cite

Non-stationarities are ubiquitous in EEG signals. They are especially apparent in the use of EEG-based brain-computer interfaces (BCIs): (a) in the differences between the initial calibration measurement and the online operation of a BCI, or (b) caused by changes in the subject's brain processes during an experiment (e.g. due to fatigue, change of task involvement, etc). In this paper, we quantify for the first time such systematic evidence of statistical differences in data recorded during offline and online sessions. Furthermore, we propose novel techniques of investigating and visualizing data distributions, which are particularly useful for the analysis of (non-)stationarities. Our study shows that the brain signals used for control can change substantially from the offline calibration sessions to online control, and also within a single session. In addition to this general characterization of the signals, we propose several adaptive classification schemes and study their performance on data recorded during online experiments. An encouraging result of our study is that surprisingly simple adaptive methods in combination with an offline feature selection scheme can significantly increase BCI performance.

show abstract

“…The data used in this work came from a public dataset for a 3-class identification problem from the BCI Competition (Millan, 2004). This dataset contains data from 3 normal subjects during 4 non-feedback sessions.…”

Section: The Eeg Signal and The Features Selectionmentioning

confidence: 99%

“…This work uses the dataset V of the BCI Competition III (Millan, 2004). The result of the competition is presented on (Blankertz et al, 2006;Blankertz, 2005).…”

Section: Introductionmentioning

confidence: 99%

A self-organizing maps classifier structure for brain computer interfaces

Bueno

Bastos

2015

Res. Biomed. Eng.

View full text Add to dashboard Cite

Introduction: Brain Computer Interfaces provide an alternative communication path to severe paralyzed people and uses electrical signals related to brain activity in order to identify the user's intention. In this paper a classifier based on a Self-Organizing Map is introduced. Methods: Electroencephalography signal is used on this work as a source for the user's intention. This signal represents the brain activity and is processed in order to extract the frequency features presented to the classifier, which uses a Self-Organizing Map and a series of probability masks in order to identify the correct class. Results: The proposed structure was evaluated using a dataset of Electroencephalography with three mental tasks. The system was able to identify the different states of the users intention with an accuracy of 71.21% for a three-class problem using only 25 neurons for one of the users. Conclusion: The classifier proposed in this paper has an accuracy that is around the value of similar works in the literature, using the same data, but using a small time window for the classification, meaning the system can have a better time response for the user.

show abstract

On the need for on-line learning in brain-computer interfaces

Cited by 145 publications

References 19 publications

Non-negative matrix factorization with α-divergence

Non-negative matrix factorization with α-divergence

Towards adaptive classification for BCI

A self-organizing maps classifier structure for brain computer interfaces

Contact Info

Product

Resources

About