Statistical methods for predicting individual movement patterns
Jenifer R. McClary · Digital Collections of Colorado (Colorado State University) · 2018
One of the most potentially useful methods for truly understanding our habits is by observing the data captured by our cellular phones.In 2009, Nokia launched the Lausanne Data Collection Campaign, which was designed to collect behavioral data from 170 volunteers through data collection software on contributed phones.In particular, GPS and WiFi connection data depicted the location of volunteers for the duration of their participation in the study.Through statistical analysis of the past location data for an individual over a prescribed duration of time, one can potentially predict the individuals future movement with desirable degrees of confidence.This research investigates a methodology for conducting predictive analysis of individual movement based on past known movements through analysis of several known time series statistical methods.Based on this analysis, I provide recommendations for application of this method to other datasets.