Skip to main content
. Author manuscript; available in PMC: 2017 Nov 1.
Published in final edited form as: Ergonomics. 2016 Mar 2;59(11):1514–1525. doi: 10.1080/00140139.2016.1146347

Algorithm 2.

Feature Vector Training Algorithm

Training Phase
inputs: feature vectors, state labels (Get and Put)

  1. Nomalize all features into zero mean and unit variance, and store the normalization factors for testing phase.

  2. Train the k-nearest neighborhood classifier by using the chosen feature vectors and the MVTA labels of the training data.

Testing Phase
inputs: frame number, feature vectors
output: states (Get or Put) for each frame

  1. Normalize all features with the normalizartion factors from training phase.

  2. Input the feature vectors into the classifier trained in the training phase frame by frame to classify each frame to be Get or Put.

  3. The estimated duty cycle is computed by (# of Put)/((# of Get) + (# of Put))