Skip to main content

Table 6 Comparative analysis for High-level activity recognition

From: CAPHAR: context-aware personalized human activity recognition using associative learning in smart environments

Subjects

RF

HMM

ECOC (AB)

LSTM

AL

# of Act

Subject 1

50.00%

51.02%

56.25%

62.96%

64.91%

12

Subject 2

48.96%

56.14%

59.60%

64.87%

69.39%

13

Subject 3

51.25%

49.78%

55.27%

59.95%

67.71%

13

Subject 4

53.56%

51.87%

61.84%

67.14%

73.91%

12

Subject 5

39.78%

34.25%

45.52%

46.20%

54.52%

8

Subject 6

52.37%

58.96%

67.30%

66.52%

72.00%

11

Subject 7

51.90%

59.00%

66.67%

65.36%

72.54%

10

  1. RF, Random Forest; HMM, Hidden Markov Model; ECOC, Error Correcting Output Codes; AB, AdaBoost; LSTM, Long-Short Term Memory Networks; AL, Associative Learning; # of Act, Number of Activities