Combining Forecasts

2008 ◽  
pp. 272-310
Keyword(s):  
2019 ◽  
Vol 16 (5) ◽  
pp. 4-14
Author(s):  
A. A. Surkov

Purpose of the study. The aim of this work is to consider the possibility of using expert information when combining forecasts as an additional factor in improving the accuracy of economic forecasting. Using the methodology of combining forecasts is increasingly found in the domestic practice of economic forecasting. Most researchers agree that combining forecasts improves forecasting accuracy by using all available information about the process under study, which is included in individual forecasting methods.  Today, there are many methods for constructing weighting factors when combining forecasts, but all of them are primarily based on the use of only statistical information about the process under study. But economic forecasting cannot be linear in its dynamics, many external factors constantly affect the forecasted process, and some internal ones may not be affected by the methods used. In this case, it is necessary to attract expert information or external information about the forecast obtained in order to increase its accuracy and adjust the further development of the economic process. This is especially true today, during the period of digitalization of the economy and the increasing influence of social and political factors on the dynamics of economic phenomena.  Materials and methods. For this purpose, methods of constructing integral indicators based on expert information or directly using such information at the stage of constructing a joint forecast can be directly used to make adjustments to the resulting combined forecast. Some of these approaches are already used in foreign practice of economic forecasting, while in domestic practice they are still little known. One of such approaches may be the use of the pairwise preference method or the application of Fishburn formulas for ranking particular forecasting methods by accuracy. The approaches considered in this work can be used as tools for constructing weight coefficients or as a correction of the obtained forecasting results.  Results. As a result of this article, attempts have been made to propose possible methods for combining forecasts using expert information, a summary table has been compiled with an assessment of one or another method of combining forecasts, and conclusions are drawn on the appropriateness of their application in practice. Such a table will make it possible to better understand the direction of attracting expert information to combine forecasts and choose the most suitable approach for further use in practice.  Conclusion. Combining forecasts has long established itself as an effective method for increasing forecast accuracy. This technique cannot degrade the result, in most cases increasing accuracy. The use of expert information in combining forecasts is the next step in improving this technique and requires a separate further practical study of possible tools for attracting expert information to the pool.  


2021 ◽  
pp. 109634802110478
Author(s):  
Yi-Chung Hu ◽  
Geng Wu ◽  
Peng Jiang

Accurately forecasting the demand for tourism can help governments formulate industrial policies and guide the business sector in investment planning. Combining forecasts can improve the accuracy of forecasting the demand for tourism, but limited work has been devoted to developing such combinations. This article addresses two significant issues in this context. First, the linear combination is the commonly used method of combining tourism forecasts. However, additive techniques unreasonably ignore interactions among the inputs. Second, the available data often do not adhere to specific statistical assumptions. Grey prediction has thus drawn attention because it does not require that the data follow any statistical distribution. This study proposes a nonadditive combination method by using the fuzzy integral to integrate single-model forecasts obtained from individual grey prediction models. Using China and Taiwan tourism demand as empirical cases, the results show that the proposed method outperforms the other combined methods considered here.


2008 ◽  
pp. 2792-2797
Author(s):  
Chi Kin Chan

The traditional approach to forecasting involves choosing the forecasting method judged most appropriate of the available methods and applying it to some specific situations. The choice of a method depends upon the characteristics of the series and the type of application. The rationale behind such an approach is the notion that a “best” method exists and can be identified. Further that the “best” method for the past will continue to be the best for the future. An alternative to the traditional approach is to aggregate information from different forecasting methods by aggregating forecasts. This eliminates the problem of having to select a single method and rely exclusively on its forecasts.


Author(s):  
Ingrida Radziukyniene ◽  
Petros Xanthopoulos ◽  
Panos M. Pardalos
Keyword(s):  

1989 ◽  
Vol 20 (4) ◽  
pp. 178-183
Author(s):  
C. B. Calitz ◽  
E. V.D.M. Smit

In the literature on forecasting, consensus has been reached about improved forecasting accuracy brought about by the combination of two or more forecasts for a given variable. No consensus, however, exists about the exact way in which the various forecasts in the combination should be weighed. The evidence points towards simple weighing schemes. The present study utilizes South African macro-economic forecasts published by seven forecasters on eight variables to evaluate the benefits to be gained from combining forecasts and to evaluate the relative accuracy of a number of combination schemes. The results confirm the current views on the combination of forecasts in so far as combining forecasts have led to increased accuracy in forecasting. It further confirms the viewpoint that a simple or weighted arithmetic average of individual forecasts seems to be acceptable as instruments for combining forecasts.


Sign in / Sign up

Export Citation Format

Share Document