Weighted Combined Forecasting Methods

The thought process around weighted combined forecasting, also known as composite forecasting, is not new. The topic has been studied for years, and empirical evidence indicates that the combination of forecast methods tends to outperform most single forecast methods. The earliest research on the value of using a weighted combined forecast was conducted by asking people to estimate distance, volume, or weight. The most common form of combined weighted forecasting is a group of judges who assign scores for athletes at sporting events, such as boxing. By averaging the scores of any three judges at a given boxing event, the obtained combined scores produce a better chance of being correct regarding who the actual winner should be. In 1969, James Bates and Clive Granger suggested that if the objective is to produce accurate forecasts, then the analyst should attempt to combine forecasts.1 Those combined forecasts should be preferred over the individual forecasts used to generate the composite. The idea behind combining forecasts is that it is better to average forecasts in hope that, by so doing, the biases among the methods and/or forecasters will compensate for one another. As a result, forecasts developed using different methods are expected to be more useful in cases of high uncertainty. To expand on this notion of uncertainty, the use of a combination of methods would seem especially relevant for long-range forecasting, where uncertainty ...

Get Demand-Driven Forecasting: A Structured Approach to Forecasting, 2nd Edition now with the O’Reilly learning platform.

O’Reilly members experience live online training, plus books, videos, and digital content from nearly 200 publishers.