Skip to main content

Table 2 Offline and real-time results

From: BioPatRec: A modular research platform for the control of artificial limbs based on pattern recognition algorithms

 

Offline

 

Real-time

 

Classification of

Training

Accuracy

 

Selection

Completion

Completion

Accuracy

 

testing sets (ms)

time (s)

(%)

 

time (s)

time (s)

rate (%)

(%)

LDA

1.03 (±0.018)

0.125 (±0.002)

92.1 (±4)

 

0.622 (±0.24)

1.86 (±0.31)

87.3 (±11)

67.1 (±10)

MLP

0.58 (±0.003)

164.1 (±52.06)

91.2 (±5)

 

0.807 (±0.27)

2.18 (±0.32)

75.8 (±13)

60.9 (±8.8)

RFN

1.49 (±0.012)

0.552 (±0.007)

83.8 (±9)

 

0.627 (±0.22)

1.89 (±0.30)

78.0 (±12)

67.4 (±10)

  1. Performance summary of each classifier in predicting 11 movements in 17 subjects (offline) and 15 subjects (real-time).