Consensus Forecasting Using Relative Error Weights
Conway Lackman
(Associate Professor in the Palumbo School of Business at
Duquesne University, Pittsburgh, Pennysylvania, USA)
Charles Brandon
(Associate Professor at Rollins College, Winter Park,
Florida, USA)
1078
Abstract
Employs weights based on relative errors produced by two or more models to formulate a consensus forecast. This method better captures changing patterns through time and sharply reduces size of forecast error.
Keywords
Citation
Lackman, C. and Brandon, C. (1994), "Consensus Forecasting Using Relative Error Weights", Marketing Intelligence & Planning, Vol. 12 No. 1, pp. 37-41. https://doi.org/10.1108/02634509410052649
Publisher
:MCB UP Ltd
Copyright © 1994, MCB UP Limited