Skip to main content
. 2021 Jan 28;14:629572. doi: 10.3389/fnins.2020.629572

Table 3.

The average accuracy of low-precision sessions (≤ 60%) for transfer learning algorithms (i.e., LRTL, wLRTL) and baseline algorithm (i.e., LR) under the inter-subject condition.

Methods S1 S2 S3 S4 S5 S6 S7
LR (%) 56.5 52.9 52.8 55.6 55.6 55.1 54.2
LRTL (%) 60.2 57.3 58.9 58.3 58.3 58.9 59.5
wLRTL (%) 59.9 57.6 59.4 58.9 58.7 59.4 60.2

The accuracies of transfer learning approaches were higher than that of baseline approach.