Beyond Where to How: A Machine Learning Approach for Sensing Mobility Contexts Using Smartphone Sensors

R.E. Guinness

Abstract:	As Global Navigation Satellite System (GNSS) technologies continue to mature, it is important to find new application areas where GNSS can be utilized in daily life. This paper presents results of research on the use of smartphone sensors (namely GPS and accelerometers), geospatial information (points of interest, such as bus stops and train stations), and machine learning (ML) to sense mobility contexts. Our goal is to develop techniques to continuously and automatically detect a smartphone user’s mobility activities, including walking, running, driving, and using a bus or train, in real-time or near real-time (<5 seconds). Automatic detection of such contexts will enable many new applications in the field of mobile computing, such as automatic route/timetable planning, carbon footprint monitoring, automatic call routing/messaging (especially while driving), or personal health monitoring. In this study, we used only two smartphone sensors, a GPS receiver and a three-axis accelerometer, in order to reduce complexity. To obtain contextual geospatial information, such as the locations of bus stops and train stations, we used two freely-available web services. We investigated a wide range of supervised learning techniques for classification, including decision trees (DT), support vector machines (SVM), naïve Bayes classifiers (NB), Bayesian Networks (BN), logistic regression (LR), artificial neural networks (ANN), and several instance-based classifiers (KStar, LWL, and IBk). Applying ten-fold cross-validation, the overall performance of these ML techniques, expressed as the correct classification rate, is as follows: 96.5% for DT (best out of multiple variants), 90.9% for BN, 87.2% for ANN, 83.4% for LR, 81.5% for NB, 80.2% for SVM, and for instance-based classifiers, 96.6%, 95.6%, and 80.3% for LWL, KStar, and IBk, respectively. Lastly, we measured the computational complexity of the classifiers, in terms of central processing unit (CPU) time needed for classification, which is important in mobile computing applications. As a result, the classifiers can be ranked from lowest to highest complexity (i.e. computational cost) as follows: SVM, ANN, LR, BN, DT, NB, IBk, LWL, and KStar. The instance-based classifiers take considerably more computational time than the noninstance-based classifiers, whereas the slowest noninstance-based classifier (NB) required about five times the amount of CPU time as the fastest classifier (SVM). The above results suggest that DT algorithms are excellent candidates for detecting mobility contexts in smartphones, both in terms of performance and computational complexity.
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	2868 - 2879
Cite this article:	Guinness, R.E., "Beyond Where to How: A Machine Learning Approach for Sensing Mobility Contexts Using Smartphone Sensors," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 2868-2879.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In