Skip to main content

Table 5 Classification accuracy of each validation method shown in Tables 2, 3 and 4 using different classifiers

From: In-lab versus at-home activity recognition in ambulatory subjects with incomplete spinal cord injury

Validation method

SVM

Naive bayes

Logisitic regression

kNN

Decision tree

Within-subject, in-lab activity

91.2%

91.9%

90.8%

86.1%

89.6%

Within-subject, train in-lab, test at-home

54.6%

45.5%

47.7%

54.2%

49.0%

Within-subject, at-home activity

85.6%

79.4%

85.1%

79.6%

82.1%

  1. The highest accuracy classifier for each validation method is indicated in bold