Optimization conditions of OCSVM for erroneous GPS data filtering

Abstract

The topics on human mobility model have long been researched by various academic and industrial fields. It has been proven that human mobility has specific patterns and can be predicted up to the probability of 93%, since the mobility of a person cannot be random while peoples have their own frequent visiting places such as home, office, haunt restaurants, and so on. The positioning data of a human can be obtained by GPS or similar positioning system, however, it contains inherited environmental errors. In this paper we will present filtering method of erroneous GPS data of human mobility. With the use of One Class Support Vector Machine (OCSVM), we adapted Radial Basis Function (RBF) as kernel function. Experimental values of the critical parameter γ for RBF has been found for optimal filtering. © 2011 Springer-Verlag.