EMAIL: PASSWORD:
Front Office
UPT. PERPUSTAKAAN
Institut Teknologi Sepuluh Nopember Surabaya


Kampus ITS Sukolilo - Surabaya 60111

Phone : 031-5921733 , 5923623
Fax : 031-5937774
E-mail : libits@its.ac.id
Website : http://library.its.ac.id

Support (Customer Service) :
timit_perpus@its.ac.id




Welcome..guys!

Have a problem with your access?
Please, contact our technical support below:
LIVE SUPPORT


Moh. Fandika Aqsa


Davi Wahyuni


Tondo Indra Nyata


Anis Wulandari


Ansi Aflacha




ITS » Master Theses » Teknik Informatika - S2
Posted by aguss at 19/07/2007 11:29:15  •  47631 Views


PERBANDINGAN METODE NEURAL NETWORKS DAN METODE ARIMA UNTUK PERAMALAN DATA TIME SERIES

COMPARATIVE METHODS OF NEURAL NETWORKS AND ARIMA FOR TIME SERIES FORECASTING

Author :
Saikhu, Ahmad 




ABSTRAK

Pemodelan data time series digunakan untuk proses peramalan karakteristik tertentu melakukan pengendalian ataupun untuk mengenali perilaku sistem sehingga model harus mempunyai kinerja yang tinggi. Pada umumnya metode statistik untuk peramalan data time series stasioner dan non stasioner kurang akurat dan tidak lengkap dalam mendeskripsikan perilaku data sehingga hasil peramalan menyimpang dari nilai aktual. Berkaitan dengan hal tersebut Tesis ini mengusulkan pengembangan Neural Networks untuk peramalan. Pengembangan ini meliputi teknik identifikasi tipe data time series dengan uji Dickey Fuller teknik diferensi penentuan jumlah input berdasarkan PACF dan model aditif Neural Networks. Diharapkan pengembangan ini akan mempercepat proses dan meningkatkan kinerja model. Studi empiris pada empat tipe data time series menunjukkan bahwa akurasi peramalan dari Metode Neural Networks relatif lebih baik dibandingkan dengan ARIMA. Pemodelan Aditif akan meningkatkan kinerja dari model Neural Networks tetapi tidak selalu lebih baik dalam peramalan beberapa periode ke depan dibandingkan dengan Nerual Networks tanpa Aditif. Pada tipe data non stasioner penerapan teknik diferensi akan meningkatkan kinerja model dan akurasi peramalan.


ABSTRACT

Time Series data modelling is used to some characteristic forecasting process doing control or recognize behaviour system so that the model must have a high works performance. Commontly the statistics methods for stationer and non stationer time series forecasting are less adequate and not complete. In descripting data behaviour so that the result of forecasting deviate from the actual value. Related to that problem this thesis propose to Developt Neural Networks for forecasting. This development include the identification technique type of time series data using Dickey Fuller Test differencing technique determining input number based on PACF and additive model Neural Networks. Hopefully development will accelerate the process and increase the model works performance. Empirical Study of four type time series data show that forecast accuration from Neural Networks Method is better than ARIMA method. Additive Modelling will increase works performance of Neural Networks model but it is not always better in future periods forecasting compare with Neural Networks without Additve. In non stationer data type implementation of difference technique will increase works performance model and forecasting accuracy.



KeywordsPeramalan ; Data time series ; metode arima
 
Subject:  Perangkat lunak komputer
Contributor
  1. Drs.Ec. Ir. Riyanarto Sarno, M.Sc, Ph.D.
Date Create: 19/07/2007
Type: Text
Format: pdf
Language: Indonesian
Identifier: ITS-Master-3100002014673
Collection ID: 3100002014673
Call Number: 006.3 Sai p


Coverage
ITS Community only




[ Download - Summary ]

ITS-Master-3100002014673-1608.pdf




 Similar Document...




! ATTENTION !

To facilitate the activation process, please fill out the member application form correctly and completely

Registration activation of our members will process up to max 24 hours (confirm by email). Please wait patiently

POLLING

Bagaimana pendapat Anda tentang layanan repository kami ?

Bagus Sekali
Baik
Biasa
Jelek
Mengecewakan





You are connected from 35.173.47.43
using CCBot/2.0 (https://commoncrawl.org/faq/)



Copyright © ITS Library 2006 - 2019 - All rights reserved.
Dublin Core Metadata Initiative and OpenArchives Compatible
Developed by Hassan