The Effect of Stepsize on Prediction Accuracy in Time Series Models
Dr. Yew-Meng Koh, Mathematics
Prediction accuracy of the response variable at future time points in a time series is a function of many factors, one of which is the number of available observations. We look at how adversely larger stepsizes (which are related to the number of available observations) affect prediction accuracy. We also consider the effect of using information from additional predictor variables on prediction accuracy as a function of stepsize.
A recommended citation will become available once a downloadable file has been added to this entry.