Statistical estimation of origin and destination points of trip using plurality of types of data sources

Abstract

A method of predicting the origin and destination points of an unknown trip using a computer includes receiving an input of second marker information including the type and position of a known marker included in a second region; generating a second feature vector at each spot included in the second region on the basis of the second marker information; and predicting the probability that the respective spots included in the second region are the origin and destination points on the basis of a prediction model, which is acquired based on first marker information including the type and position of a known marker included in a first region and information on the known origin and destination points included in the first region, and the second feature vector.

Dependencies are inferred from the claim text.