Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Multi-disciplinary Studies | en_US |
| dc.contributor | Department of Applied Mathematics | en_US |
| dc.creator | Wong, Chi-shing | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/3005 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | - |
| dc.rights | All rights reserved | en_US |
| dc.title | A study of combination of forecasts | en_US |
| dcterms.abstract | The objective of this project is to investigate the performance of conventional combined forecast methods and proposed new combined forecast methods. The new methods proposed use a balance factor calculated from three different ways (denoted by B1, B2 and B3) to compensate the bias presented in the individual forecast. In addition, skewness factor was proposed to handle the error pattern for weight computation. Twenty time series were collected and used for the investigation. Four individual forecasts were selected for combination. They were (1) Box-Jenkins method, (2) Holt-Winters method, (3) adaptive-response-rate simple exponential smoothing method and (4) double moving average method. Eight conventional combined forecast methods were selected from the past researches. With different settings of parameters, these eight methods expanded to 52 methods for investigation. Also the proposed new methods using balance factor and skewness factor provided 21 methods for investigation. According to the findings of this study, the best conventional methods provided a U-statistic and MAPE of 0.877 and 5.60% respectively. However, it did not out perform the ARIMA model which U-statistic and MAPE were 0.875 and 5.41% respectively. The proposed skewness factor was found successful to provide a slightly improved performance. However, it is not recommended to apply for forecasting work if cost and effectiveness are considered. The proposed balance factor from method B1 were found successful to provide better performance than conventional methods on a condition that this method applies only to the individual forecasts in which autocorrelation of residuals is present. The best U-statistic and MAPE of this method were 0.858 and 5.16% respectively. The proposed balance factor from method B2 was not successful. The proposed balance factor from method B3 was found successful and the best U-statistic and MAPE were 0.867 and 5.55%. However, unlike the case of method B1, the autocorrelation of residuals does not have significant effect on its application. | en_US |
| dcterms.extent | vi, 118 : ill. ; 30 cm | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2000 | en_US |
| dcterms.educationalLevel | All Master | en_US |
| dcterms.educationalLevel | M.Sc. | en_US |
| dcterms.LCSH | Forecasting -- Mathematical models | en_US |
| dcterms.LCSH | Hong Kong Polytechnic University -- Dissertations | en_US |
| dcterms.accessRights | restricted access | en_US |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| b15354349.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 3.66 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/3005