Limits...
Collaborative brain-computer interface for aiding decision-making.

Poli R, Valeriani D, Cinel C - PLoS ONE (2014)

Bottom Line: We then built a composite neuro-behavioural feature which optimally combines the two measures.For group decisions, we uses a majority rule and three rules which weigh the decisions of each observer based on response times and our neural and neuro-behavioural features.Results indicate that the integration of behavioural responses and neural features can significantly improve accuracy when compared with the majority rule.

View Article: PubMed Central - PubMed

Affiliation: Brain-Computer Interfaces Lab, School of Computer Science and Electronic Engineering, University of Essex, Colchester, United Kingdom.

ABSTRACT
We look at the possibility of integrating the percepts from multiple non-communicating observers as a means of achieving better joint perception and better group decisions. Our approach involves the combination of a brain-computer interface with human behavioural responses. To test ideas in controlled conditions, we asked observers to perform a simple matching task involving the rapid sequential presentation of pairs of visual patterns and the subsequent decision as whether the two patterns in a pair were the same or different. We recorded the response times of observers as well as a neural feature which predicts incorrect decisions and, thus, indirectly indicates the confidence of the decisions made by the observers. We then built a composite neuro-behavioural feature which optimally combines the two measures. For group decisions, we uses a majority rule and three rules which weigh the decisions of each observer based on response times and our neural and neuro-behavioural features. Results indicate that the integration of behavioural responses and neural features can significantly improve accuracy when compared with the majority rule. An analysis of event-related potentials indicates that substantial differences are present in the proximity of the response for correct and incorrect trials, further corroborating the idea of using hybrids of brain-computer interfaces and traditional strategies for improving decision making.

Show MeSH
Average percentage of errors vs group size and number of cross-validation folds for group decisions made with the RTnf-based method.As can be seen from the overlapping error bars (representing the standard error of the mean) and extensive statistical comparisons (see text), performance depends very little on the particular choice of the number of folds used for cross-validation.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4114490&req=5

pone-0102693-g008: Average percentage of errors vs group size and number of cross-validation folds for group decisions made with the RTnf-based method.As can be seen from the overlapping error bars (representing the standard error of the mean) and extensive statistical comparisons (see text), performance depends very little on the particular choice of the number of folds used for cross-validation.

Mentions: We should note that the results obtained by using nf and RTnf to measure confidence are very little influenced by the number of folds chosen for cross-validation (while, of course, the results of majority and the RT-based method are exactly the same for any choice of folds as no learning process takes place in such methods). To illustrate this, in Figure 8 we report the error rates for the RTnf-based method as a function of group size and number of folds. The error bars in the plots represent the standard error of the mean. A statistical comparison of the performance obtained with different numbers of folds using the Wilcoxon exact test with Bonferroni correction showed that in only 13.8% of the 550 comparisons required by a full analysis (with 11 cross-validations, there are pairwise data-set comparisons for each group size) differences were statistically significant. Also, for most group sizes the differences are very small. This suggests that the case of 16 folds on which we focused in most of the paper is reasonably representative.


Collaborative brain-computer interface for aiding decision-making.

Poli R, Valeriani D, Cinel C - PLoS ONE (2014)

Average percentage of errors vs group size and number of cross-validation folds for group decisions made with the RTnf-based method.As can be seen from the overlapping error bars (representing the standard error of the mean) and extensive statistical comparisons (see text), performance depends very little on the particular choice of the number of folds used for cross-validation.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4114490&req=5

pone-0102693-g008: Average percentage of errors vs group size and number of cross-validation folds for group decisions made with the RTnf-based method.As can be seen from the overlapping error bars (representing the standard error of the mean) and extensive statistical comparisons (see text), performance depends very little on the particular choice of the number of folds used for cross-validation.
Mentions: We should note that the results obtained by using nf and RTnf to measure confidence are very little influenced by the number of folds chosen for cross-validation (while, of course, the results of majority and the RT-based method are exactly the same for any choice of folds as no learning process takes place in such methods). To illustrate this, in Figure 8 we report the error rates for the RTnf-based method as a function of group size and number of folds. The error bars in the plots represent the standard error of the mean. A statistical comparison of the performance obtained with different numbers of folds using the Wilcoxon exact test with Bonferroni correction showed that in only 13.8% of the 550 comparisons required by a full analysis (with 11 cross-validations, there are pairwise data-set comparisons for each group size) differences were statistically significant. Also, for most group sizes the differences are very small. This suggests that the case of 16 folds on which we focused in most of the paper is reasonably representative.

Bottom Line: We then built a composite neuro-behavioural feature which optimally combines the two measures.For group decisions, we uses a majority rule and three rules which weigh the decisions of each observer based on response times and our neural and neuro-behavioural features.Results indicate that the integration of behavioural responses and neural features can significantly improve accuracy when compared with the majority rule.

View Article: PubMed Central - PubMed

Affiliation: Brain-Computer Interfaces Lab, School of Computer Science and Electronic Engineering, University of Essex, Colchester, United Kingdom.

ABSTRACT
We look at the possibility of integrating the percepts from multiple non-communicating observers as a means of achieving better joint perception and better group decisions. Our approach involves the combination of a brain-computer interface with human behavioural responses. To test ideas in controlled conditions, we asked observers to perform a simple matching task involving the rapid sequential presentation of pairs of visual patterns and the subsequent decision as whether the two patterns in a pair were the same or different. We recorded the response times of observers as well as a neural feature which predicts incorrect decisions and, thus, indirectly indicates the confidence of the decisions made by the observers. We then built a composite neuro-behavioural feature which optimally combines the two measures. For group decisions, we uses a majority rule and three rules which weigh the decisions of each observer based on response times and our neural and neuro-behavioural features. Results indicate that the integration of behavioural responses and neural features can significantly improve accuracy when compared with the majority rule. An analysis of event-related potentials indicates that substantial differences are present in the proximity of the response for correct and incorrect trials, further corroborating the idea of using hybrids of brain-computer interfaces and traditional strategies for improving decision making.

Show MeSH