Table 4.

Performance of the committee of 15 greedy classifiers trained using only fixations on button as target fixations (Trainset 1).

Sbj.	Test on fixations with 500 ms threshold				Test on fixations with 1000 ms threshold
	Specificity	Sensitivity			Sensitivity
		Button	Ball	Free cell	Button	Ball	Free cell
1	0.92 ± 0.04	0.31 ± 0.08	0.16	0.14	0.23	0.24	0.13
2	0.92 ± 0.03	0.40 ± 0.16	0.17	0.12	0.17	0.17	0.16
3	1.00 ± 0.00	0.20 ± 0.06	0.06	0.02	0.07	0.05	0.06
4	0.83 ± 0.04	0.44 ± 0.08	0.28	0.18	0.36	0.27	0.17
5	0.85 ± 0.06	0.45 ± 0.10	0.20	0.17	0.31	0.19	0.17
6	0.90 ± 0.06	0.39 ± 0.12	0.14	0.14	0.20	0.13	0.11
7	0.86 ± 0.06	0.47 ± 0.16	0.33	0.25	0.34	0.31	0.27
8	0.95 ± 0.04	0.41 ± 0.06	0.17	0.17	0.35	0.13	0.07
Mean	0.90	0.38	0.19	0.15	0.25	0.19	0.14
Std.	0.07	0.14	0.08	0.07	0.10	0.08	0.07

For training, recordings with 500 ms dwell threshold were used. On these recordings, 5-fold cross-validation was used to estimate, as M ± SD over folds, sensitivity for Button fixations and specificity in the 500 ms dwell threshold data. For each fold, the classifier was trained on 80% and tested on 20% of the set. In the case of testing on Ball and Free cell fixations, as well as for all 1000 ms dwell threshold data, the train and test sets were not overlapped and, therefore, the classifier was applied to all test data.