Table 10 Performance analysis of ML models on Case-3 within 800–1000 m distance employing 4 sensors.

From: A cost effective real time rail track monitoring system leveraging multi sensor fusion and multi objective optimization

Types of ML algorithm

Model accuracy (%)

Precision

Recall

F1-score

Memory consumption (kB)

k-Nearest neighbors

98.66

91.74

93.51

90.83

2249.82

Decision tree

95.55

89.93

88.72

90.14

162.16

Support vector machine

98.85

90.37

92.87

93.63

127.74

Deep neural network68

92.54

93.76

91.43

90.33

4362.15

Bonsai69

97.14

94.07

91.19

93.81

0.552

Back to article page