Alteration of box-jenkins methodology by implementing genetic algorithm method
A time series is a set of values sequentially observed through time. The Box-Jenkins methodology is a systematic method of identifying, fitting, checking and using integrated autoregressive moving average time series model for forecasting. Box-Jenkins method is an appropriate for a medium to a long...
Saved in:
Main Authors: | , , |
---|---|
Format: | Article |
Published: |
American Institute of Physics Inc.
2015
|
Subjects: | |
Online Access: | http://eprints.utm.my/id/eprint/59132/ http://www.dx.doi.org/10.1063/1.4907522 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | A time series is a set of values sequentially observed through time. The Box-Jenkins methodology is a systematic method of identifying, fitting, checking and using integrated autoregressive moving average time series model for forecasting. Box-Jenkins method is an appropriate for a medium to a long length (at least 50) time series data observation. When modeling a medium to a long length (at least 50), the difficulty arose in choosing the accurate order of model identification level and to discover the right parameter estimation. This presents the development of Genetic Algorithm heuristic method in solving the identification and estimation models problems in Box-Jenkins. Data on International Tourist arrivals to Malaysia were used to illustrate the effectiveness of this proposed method. The forecast results that generated from this proposed model outperformed single traditional Box-Jenkins model. |
---|