Table 2 Experimental results on the short-time emotional response using NARI point-process models implementing quadratic and cubic autoregressive functions. For each subject, the accuracy of SVM classification is expressed in percentage considering the arousal, valence and self-reported emotion classification cases. The accuracy on the 4-class emotion classification case is shown as “Emotion on Arousal-Valence Plane”

From: Revealing Real-Time Emotional Responses: a Personalized Assessment based on Heartbeat Dynamics

Subjects

Emotion on Arousal-Valence Plane

Arousal

Valence

Emotion Self-Report

1

75.86

84.15

75.31

70.00

2

78.19

80.77

83.33

68.24

3

77.57

87.80

70.37

63.86

4

79.31

85.37

77.78

67.05

5

67.24

84.15

66.67

61.43

6

89.47

86.42

83.75

66.67

7

81.48

91.67

76.62

72.23

8

73.68

87.65

76.25

64.41

9

84.21

69.14

73.75

55.00

10

67.86

82.50

58.97

60.26

11

75.44

74.07

75.00

70.00

12

87.72

83.95

80.00

71.08

13

67.24

76.83

75.31

70.89

14

82.76

82.93

81.48

82.67

15

81.03

81.71

72.84

69.57

16

75.86

69.51

75.31

63.38

17

85.19

90.91

81.82

82.28

18

84.48

76.83

87.65

73.26

19

77.59

74.39

79.01

63.16

20

82.76

82.93

76.54

67.05

21

75.86

85.37

81.48

68.06

22

77.59

89.02

74.07

70.27

23

94.83

80.49

90.12

86.15

24

87.93

84.15

80.25

75.00

25

84.21

75.31

68.75

53.33

26

76.36

75.95

67.94

54.67

27

70.69

80.49

77.78

60.92

28

71.93

77.78

76.25

71.01

29

84.21

83.95

78.75

66.25

30

80.36

68.75

79.75

65.12