APRIL 18-21, 2017
Previous Events

Outlier Identification of Plantar Pressure Data

  • Chen-Hsiang Liao
    National Taipei University of Technology
  • Tung-Ju Hsieh
    National Taipei University of Technology


Plantar pressure data is an important foundation for sports studies. It includes 20 parameters from the index arch, left foot, and right foot plantar pressure. It causes data observation inconvenient because of complicated property and outlier. We developed a database platform for users to import plantar pressure to the database. We used outlier identification to help sports researchers studying plantar pressure data. We used plantar pressure devisee to collect plantar pressure data sets obtained from different groups of college athletes. The data sets contain a total of 1,006 participants from 26 different sport teams. In order to research foot symptoms, we propose using visualization of plantar pressure data sets for comparative visualization using parallel coordinate. Average values and standard deviation of data sets are computed. We look for outliers in the data sets using a range of statistical values such as standard deviation. It can help sports researchers to identify outliers who have high risk of symptoms. We studied tennis, table tennis, and badminton. Sports researchers can use comparative visualization to compare difference data sets. It can save sports researchers time in operate plantar data sets and help them to make observation between different sports. Further, they can diagnose foot symptoms from the clustering in the data sets.