Skip to Main Content
The objective of this study was to compare base-level and meta-level classifiers on the task of activity recognition. Five wireless kinematic sensors were attached to 25 subjects with each subject asked to complete a range of basic activities in a controlled laboratory setting. Subjects were then asked to carry out similar self-annotated activities in a random order and in an unsupervised environment. A combination of time-domain and frequency-domain features were calculated using a sliding window segmentation technique. A reduced feature set was generated using a wrapper subset evaluation technique with a linear forward search.The meta-level classifier AdaBoostM1 with C4.5 Graft as its base-level classifier achieved an overall accuracy of 95%. Equal sized datasets of subject independent data and subject dependent data were used to train this classifier and it was found that high recognition rates can be achieved without the need of user specific training.