search query: @author Armstrong, J.S. / total: 6
reference: 3 / 6
| Author: | Collopy, F. Armstrong, J.S. |
| Title: | Ruel-based forecasting: Development and validation of an expert systems approach to combining time series extrapolations |
| Journal: | Management Science
1992 : OCT, VOL. 38:10; p.1394-1414 |
| Index terms: | FORECASTING TECHNIQUES TIME SERIES ARTIFICIAL INTELLIGENCE STATISTICAL METHODS |
| Language: | eng |
| Abstract: | The feasibïity of rule-based forecasting is examined. It is a procedure that applies forecasting expertise and domain knowledge to forecasts according to features of the data. A rule base was developed to make annual extrapolation forecasts for economic and demographic time series. The development of the rule of base drew upon protocol analysis of five experts on forecasting methods. It combines forecasts from four extrapolation methods: random walks, regression, Brown's linear exponential smoothing, and Holt's exponential smoothing. The improvement in accuracy of the rule-based forecasts over equally weighted combined forecasts was statistically significant. |
SCIMA