FaceOff

Prerequisites

• Python 3
• scikit-learn

Data Description

All data files end with '.csv':

• Data files starting with '0' (e.g., 0401s1general.csv) are raw data collected via several sessions to include both face touching and non face touching behaviors;
• Data files starting with 'm' (e.g., m_onset_0.1.csv) are training data; "onset_0.x" means there only contains data from the first 0.x portion of the time window (see the accompanying paper for more details);'
• Data files starting with 'p' (e.g., p3_all.csv) are testing data;

Data Processing Scripts

• experiment.py runs several experiments reported in the aforementioned paper;
• test.py contains some testing routines used by experiment.py;
• featurize.py featurizes raw data using preprocess.py and merges different sessions' data using merge.py;
• tunehyperparam.py tunes hyperparameters for random forest models.

Citation

Chen, X.A., 2020. FaceOff: Detecting Face Touching with a Wrist-Worn Accelerometer. arXiv preprint arXiv:2008.01769.

@article{chen2020faceoff,
  title={FaceOff: Detecting Face Touching with a Wrist-Worn Accelerometer},
  author={Chen, Xiang'Anthony'},
  journal={arXiv preprint arXiv:2008.01769},
  year={2020}
}