Advances in the Theta model
The Theta model created a lot of interest in academic circles due to its surprising performance in the M3-competition. However, this interest was not followed by a large number of studies, with the exception of Hyndman and Billah in 2003. The present study discusses the advances in the model that have been made in the last five years and attempts to provide further insights into the research question: “Is the Theta model just a special case of Simple Exponential Smoothing with drift (SES-d)?” If we do not use equally weighted extrapolations of two specific Theta Lines, L(T=0) and L(T=2) in the Theta model then we end up with a far more generic model than Simple Exponential Smoothing. The paper also examines the potential of an optimization version of SES-d so as to test the results of Hyndman and Billah. In contrast to their research results, the Theta model outperforms SES-d in the Quarterly-M3 and Other-M3 subsets by 0.30% and 0.36% respectively, when the Symmetric Mean Absolute Percentage Error is used to measure accuracy.
|Date of creation:||2008|
|Date of revision:|
|Contact details of provider:|| Phone: +30-2710-230128|
Web page: http://econ.uop.gr/~econ/
More information through EDIRC
When requesting a correction, please mention this item's handle: RePEc:uop:wpaper:0023. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Kleanthis Gatziolis)
If references are entirely missing, you can add them using this form.