Prediction of missing data using spatial model: application to ITS.

The need of the Intelligent Transportation System (ITS) is increasing and real-time traffic information on roads is one of the most important parts of the ITS. Rotis Inc. services real-time traffic information on roads in Seoul, Korea. In the course of real-time traffic service, in some situations traffic data are missed on some roads. New procedures are proposed based on the spatial model and the spatio-time model, for predicting missing data by specifying suitable spatial neighborhood structure. These procedures are also compared with time series models using cross-validation statistics. For the covering abstract see ITRD E114174.