Skip to main content

Table 2 Classification accuracy for each subject.

From: Single-trial classification of motor imagery differing in task complexity: a functional near-infrared spectroscopy study

Best-performing combination
Subject No. Channel Time interval Optimal feature set Classification accuracy
1 3 9-15 s Δ[O2Hb] mean, variance, skewness, kurtosis 91.7%
2 2 5-15 s Δ[O2Hb] mean, variance 79.2%
3 3 9-15 s Δ[O2Hb] variance, skewness, kurtosis 79.2%
4 2 8-14 s Δ[O2Hb] mean, variance 75.0%
5 3 9-15 s Δ[O2Hb] mean 75.0%
6 3 7-15 s Δ[O2Hb] mean, variance, skewness 91.7%
7 1 8-14 s Δ[O2Hb] skewness 70.8%
8 2 7-12 s Δ[O2Hb] mean, variance 75.0%
9 1 5-15 s Δ[O2Hb] mean, variance 83.3%
10 3 5-15 s Δ[O2Hb]variance, skewness, kurtosis 87.5%
11 3 7-15 s Δ[O2Hb]variance, kurtosis 87.5%
12 2 11-15 s Δ[O2Hb] mean, variance, skewness, kurtosis 79.2%
Overall     81.3 ± 7.0%
  1. The results are shown for the best-performing combination of one channel, a certain time interval and the optimal feature set for each subject. Classification accuracy was identified over 12 randomised trials by cross validation. Four features were used: mean Δ[O2Hb] amplitude, variance, skewness and kurtosis