LOGIN TO YOUR ACCOUNT

Username
Password
Remember Me
Or use your Academic/Social account:

CREATE AN ACCOUNT

Or use your Academic/Social account:

Congratulations!

You have just completed your registration at OpenAire.

Before you can login to the site, you will need to activate your account. An e-mail will be sent to you with the proper instructions.

Important!

Please note that this site is currently undergoing Beta testing.
Any new content you create is not guaranteed to be present to the final version of the site upon release.

Thank you for your patience,
OpenAire Dev Team.

Close This Message

CREATE AN ACCOUNT

Name:
Username:
Password:
Verify Password:
E-mail:
Verify E-mail:
*All Fields Are Required.
Please Verify You Are Human:
fbtwitterlinkedinvimeoflicker grey 14rssslideshare1
Riedel, Silvia; Gabrys, Bogdan (2007)
Publisher: IEEE Press
Languages: English
Types: Unknown
Subjects: bmm, aintel, csi
In this paper we provide experimental results and\ud extensions to our previous theoretical findings concerning the\ud combination of forecasts that have been diversified by three\ud different methods: with parameters learned at different data\ud aggregation levels, by thick modeling and by the use of different\ud forecasting methods. An approach of error variance based\ud pooling as proposed by Aiolfi and Timmermann has been compared\ud with flat combinations as well as an alternative pooling\ud approach in which we consider information about the used\ud diversification. An advantage of our approach is that it leads to\ud the generation of novel multi step multi level forecast generation\ud structures that carry out the combination in different steps of\ud pooling corresponding to the different types of diversification.\ud We describe different evolutionary approaches in order to\ud evolve the order of pooling of the diversification dimensions.\ud Extensions of such evolutions allow the generation of more\ud flexible multi level multi step combination structures containing\ud better adaptive capabilities. We could prove a significant error\ud reduction comparing results of our generated combination\ud structures with results generated with the algorithm of Aiolfi\ud and Timmermann as well as with flat combination for the\ud application of Revenue Management seasonal forecasting.

Share - Bookmark

Cite this article