APS

31st APS Annual Convention · 2019

Comparing Data Mining Methods for Prediction of Social Distance

Washington, DC · May 2019

Poster · Methodology

  • Yaqi Li
    University of Oklahoma
  • Linle Jiang
    University of Southern California
  • Hairong Song
    The University of Oklahoma

Abstract

This study was aimed to compare the accuracy of prediction on closest friendship using data mining techniques, including random forest, Bayesian network, artificial neural network, and logistic regression. Applied with the data crawled from the Facebook website, naïve Bayesian network performed the best with the accuracy as high as 91%.

Statistics and Methodology

← Poster Session XIV