Skip to main content

Table 8 An average proportion of accuracy variation explained by the regression models fit to the evaluated metrics presented for the two types of mobility sequences of varied spatio-temporal resolutions

From: Explaining human mobility predictions through a pattern matching algorithm

 

Next-place [%]

Next time-bin [%]

 

1688 m

204m

33 m

1688 m

204m

33 m

1-hour

30-minutes

ESR

57.71

61.09

56.97

38.78

27.02

35.69

38.18

33.20

SR

26.45

41.65

47.63

47.61

29.51

46.80

42.51

43.69

DR

19.34

30.95

32.06

50.57

44.87

44.28

50.21

40.92

ESR + NStationarity

47.92

45.03

47.14

52.88

42.95

SR + NStationarity

75.63

48.92

67.64

72.68

62.60

DR + NStationarity

48.91

52.28

47.99

57.46

47.10

Regularity + Stationarity

23.79

24.36

31.89

43.28

38.12

38.49

40.62

36.44

GA

55.27

55.32

56.98

75.69

51.84

61.50

64.91

57.09

GA + NStationarity

80.55

65.48

66.49

74.56

66.41

IGA

52.97

52.74

55.95

86.23

85.44

89.67

84.44

88.03

IGA + NStationarity

88.55

86.49

90.33

87.53

89.23

\(\Pi _{\max}\)

38.32

43.39

42.10

48.56

48.57

47.70

52.92

43.63

  1. ESR is an equally sparse repeatability, SR is a sparse repeatability, DR is a dense repeatability, GA is a global alignment measure, and IGA is an iterative global alignment measure. All the \(R^{2}\) values are significant at the level of p<0.001. Bold values indicate the best result for each sequence type.