Open AccessJournal Article10.1080/03610918.2012.731121

Combining Forecasts via Simulations

- 01 Jan 2014

- Communications in Statistics - Simulatio...

- Vol. 43, Iss: 6, pp 1318-1332

TL;DR: Simulation studies and applications of the proposed method to real datasets show that on average the combined forecasts are more accurate than those obtained without combining.

Abstract: Combining forecasts can be based on different data or different methods or both. In practice, many different datasets can be difficult to obtain. In this article, we propose a combining forecasts method that uses both simulated and observed time series. Several combining methods have been studied and the combined forecasts have been compared with those obtained from the model fitted to the observed time series only. Simulation studies and applications of the proposed method to real datasets show that on average the combined forecasts are more accurate than those obtained without combining.

AI Agents for this Paper

Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps

Most frequently asked questions

1. What contributions have the authors mentioned in the paper "Combining forecasts via simulations" ?

In this paper the authors propose a combining forecasts method that uses both simulated and observed time series.. Several combining methods have been studied and the combined forecasts have been compared with those obtained from the model fitted to the observed time series only.

2. What are the future works in "Combining forecasts via simulations" ?

The authors feel that the optimal length of simulated time series is certainly worth further investigation in the future.. However, a thorough analysis on this ex ante condition by using simulated time series needs to be carried in the future.. Graefe et al. ( 2010 ) pointed out that forecasts from different methods that use different data may lead to bracketing and low correlations of errors, and in such situations, the gains from combining can be substantial.. This paper suggests that the simulated time series can be used for this purpose and that combining does help in improving forecasting accuracy.

Table 1: Estimated parameter values of the base processes

Figure 4: Autocorrelation structures of the simulated and the observed HSI, FIN and C&I time series.

Figure 1: (a) Observed (darker curve) and simulated (lighter curve) time series. (b) Marginal distributions. (c) (d) auto-correlation structures. (e) QQ-plot of the observed and the simulated time series in Simulation study 1.

Table 2: RMSE and MdAPE values between forecasts and actual observed data in the simulation studies

Figure 3: Observed (darker curve) and simulated (lighter curve) HSI, FIN and C&I time series and the corresponding QQ-plots.

Table 5: Average RMSE and MdAPE values for different methods and different lengths of the simulated time series

Citations

•Journal Article•10.4018/IJEC.290292

Energy-Efficient Cloud-Integrated Sensor Network Model Based on Data Forecasting Through ARIMA

Kalyan Das, +1 more

- 01 Jan 2022

- International Journal of e-Collaboration

References

•Book

Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach

Kenneth P. Burnham, +1 more

- 19 Jun 2013

TL;DR: The second edition of this book is unique in that it focuses on methods for making formal statistical inference from all the models in an a priori set (Multi-Model Inference).

...read moreread less

45.1K

•Journal Article

Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach (2nd ed.)

Mario Peruggia

- 01 Sep 2003

- Journal of the American Statistical Asso...

TL;DR: In this age of modern era, the use of internet must be maximized, to get the on-line model selection and multi model inference a practical information theoretic approach book, as the world window, as many people suggest.

...read moreread less

2.5K

Journal Article•10.1016/0169-2070(89)90012-5

Combining forecasts: A review and annotated bibliography

Robert T. Clemen

- 01 Jan 1989

- International Journal of Forecasting

TL;DR: In this article, the authors provide a review and annotated bibliography of that literature, including contributions from the forecasting, psychology, statistics, and management science literatures, providing a guide to the literature for students and researchers and to help researchers locate contributions in specific areas, both theoretical and applied.

...read moreread less

2.3K

•Journal Article•10.1016/0169-2070(92)90008-W

Error measures for generalizing about forecasting methods: Empirical comparisons

J. Scott Armstrong, +1 more

- 01 Jun 1992

- International Journal of Forecasting

TL;DR: In this article, the authors evaluated measures for making comparisons of errors across time series and found that the median absolute error of a given method to that from the random walk forecast is not reliable, and therefore inappropriate for comparing accuracy across series.

...read moreread less

1.4K

Book Chapter•10.1093/oso/9780198522249.001.0001

Non-linear Time Series

Howell Tong

- 02 Aug 1990

TL;DR: Non-linear time series analysis is a rapidly developing subject. This book introduces some of the recent developments and the current state of research. It covers dynamical systems theory and its role in non-linear time series analysis.

...read moreread less

1.2K

...

Expand

Combining Forecasts via Simulations

Chat with Paper

AI Agents for this Paper

Most frequently asked questions

1. What contributions have the authors mentioned in the paper "Combining forecasts via simulations" ?

2. What are the future works in "Combining forecasts via simulations" ?

Figures

Citations

Energy-Efficient Cloud-Integrated Sensor Network Model Based on Data Forecasting Through ARIMA

References

Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach

Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach (2nd ed.)

Combining forecasts: A review and annotated bibliography

Error measures for generalizing about forecasting methods: Empirical comparisons

Non-linear Time Series

Related Papers (5)

Combination of Economic Forecasts: An Odds-Matrix Approach

N‐step combinations of forecasts

Combining predictive distributions for the statistical post-processing of ensemble forecasts

Calibrating probabilistic forecasts from an NWP ensemble

Ensemble solar forecasting using data-driven models with probabilistic post-processing through GAMLSS