Development and validation of a rule-based time series complexity scoring technique to support design of adaptive forecasting DSS
详细信息 查看全文
- 作者：Monica Adyaa ; 1 ; monica.adya@marquette.edu" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Edward J. Luskb ; c ; d ; 2 ; luskej@plattsburgh.edu" class="auth_mail" title="E-mail the corresponding author ; lusk@wharton.upenn.edu" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae
- 关键词：Benchmark forecasting ; Forecasting decision support systems ; Structured judgment ; Forecasting ; Time series ; Rule-based Forecasting
- 刊名：Decision Support Systems
- 出版年：2016
- 出版时间：March 2016
- 年：2016
- 卷：83
- 期：Complete
- 页码：70-82
- 全文大小：910 K
文摘
- class="listitem" id="list_l0005">
- class="label">•
A rule-based method, CST, for scoring time series complexity is developed and validated.
- class="label">•
Validations confirm that the CST meaningfully distinguishes simple from complex series as below.
- class="label">•
Forecast errors from benchmark methods for simple series are lower than for those scored as complex.
- class="label">•
CST can be integrated into FDSS to provide decision support that adapts to time series complexity.
- class="label">•
A framework for development of such an adaptive system is proposed.