Analysis of Different Classification Techniques for Two-Class Functional Near-Infrared Spectroscopy-Based Brain-Computer Interface

Table 8

Averaged values of the classification accuracies, precisions, and recalls of 3-feature combinations across all subjects.

Classifiers	S1	S2	S3	S4	S5	S6	S7	Average
LDA
Accuracy	81.24	81.76	78.41	80.54	79.68	77.88	77.78	79.6 1.5
Precision	89.9	89.21	89.66	75.09	74.21	72.37	72.53	80.4 7.8
Recall	73.94	73.98	65.79	91.33	90.09	89.88	87.48	81.8 9.5
QDA
Accuracy	95.84	96.58	93.49	95.96	95.11	94.12	95.93	95.2 1
Precision	97.26	98.67	98.67	93.79	92.34	90.8	93.85	95.1 2.9
Recall	94.41	94.1	88.19	98.71	98.64	98.8	98.66	95.9 3.7
NN
Accuracy	63.92	65.32	64.85	64.78	64.59	63.55	65.11	64.5 0.3
Precision	63.22	62.21	63.85	64.13	65.85	61.04	61.07	63.1 1.6
Recall	65.28	67.56	66.15	61.67	69.49	60.13	61.54	64.6 3.2
Naïve Bayes
Accuracy	95.58	96.34	92.77	95.53	94.72	93.47	95.55	94.8 1.2
Precision	94.45	94.22	88.46	98.73	98.35	98.64	98.68	95.9 3.6
Recall	96.98	99.12	98.95	92.22	90.89	88.1	92.29	94.1 3.9
SVM
Accuracy	95.79	95.75	94.21	95.97	95.58	94.47	94.85	95.2 0.7
Precision	93.22	99.03	98.67	93.79	92.34	90.8	93.85	94.5 2.9
Recall	99.12	94.1	88.19	98.71	98.64	98.8	98.66	96.6 3.8
ANN
Accuracy	96.48	96.49	95.78	96.87	96.4	95.97	96.41	96.3 0.3
Precision	93.9	94.4	98.1	95.05	93.26	91.59	93.71	94.3 1.9
Recall	91.9	92.8	91.55	93.64	98.39	98.46	98.67	95.1 3.0