Models of detection relationship between time series in forecasting problems

The problem of forecasting multiple time series requires detection of relationship between them. Engagement of related time series in a forecast model boosts the forecast quality. This paper introduces the convergent cross mapping (CCM) method used to detect a relationship between time series. This method estimates accuracy of reconstruction of one time series using the other series. The CCM method detects relationship between series not only in full trajectory spaces, but also in some trajectory subspaces. The computational experiment was carried out on two sets of time series: electricity consumption and air temperature, oil transportation volume and oil production volume.