METODE WAVELET MODWT UNTUK PERAMALAN DATA RUNTUN WAKTU NON MUSIMAN
| Main Authors: | , WIWIK WIYANTI, S.SI, , Prof. Drs. H. Subanar, Ph. D |
|---|---|
| Format: | Thesis NonPeerReviewed |
| Terbitan: |
[Yogyakarta] : Universitas Gadjah Mada
, 2012
|
| Subjects: | |
| Online Access: |
https://repository.ugm.ac.id/100068/ http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=56411 |
Daftar Isi:
- Predicting is an estimate of the systematic process about what is most likely to occur in the future which based on past information and the present (time series) owned to get mistake or the difference between what happens with the results of the estimation can be minimized or reduced. The method is often used in predicting numbers or data that may occur in the future is the Box-Jenkins methods. In this present study was to see those numbers will be used a method called wavelet MODWT (Maximal Overlap Discrete Wavelet Transform). MODWT used because this method is very suitable for data that have a particularly sharp trend of non-seasonal data. The result from MODWT will be obtained smooth and detail, and then smooth and detail coefficient would used to get MAR (Multiscale Autoregressive) which will be used to predict the numbers that may occur. From the empirical results in ten non seasonal time series data research found that estimated value of eight data research is the value forecast error with MODWT smaller than Box Jenkins, while the value forecast error two data is with Box Jenkins method smaller than MODWT.