Intelligent Database Systems Lab Presenter : Kung, Chien-Hao Authors : Medhdi Khashei, Mehdi Bijari 2011, ASOC A novel hybridization of artificial neural.

Intelligent Database Systems Lab

Presenter : Kung, Chien-Hao

Authors : Medhdi Khashei, Mehdi Bijari

2011, ASOC

A novel hybridization of artificial neural networks and ARIMA models

for time series forecasting


Outlines

MotivationObjectivesMethodologyExperimentsConclusionsComments


Motivation• Artificial neural networks(ANNs) are applied

to a wide range of forecasting problems.

• However, using ANNs to model linear problems

have yielded mixed results, and hence; it’s not wise

to apply ANNs blindly to any type of data.


Objectives• To overcome limitation of ANNs and yield more

general and more accurate forecasting model.


Methodology-Framework

ARIMA

ANNs

Forecast


Methodology

ARIMA

ANNs

Forecast

AR MA

ARMA

ARIMA


Methodology

ARIMA

ANNs

Forecast


Methodology

ARIMA

ANNs

Forecast


Experiments-Sunspot dataset


Experiments-Lynx dataset


Experiments-Exchange rate dataset


Experiments


Conclusions

• A novel hybridization of ANNs and ARIMA model is

proposed in order to overcome the limitation of ANNs

and yield the more general and the more accurate

forecasting model.


Comments• Advantages– The experiment is rich.

• Applications– Time series forecasting.

Intelligent Database Systems Lab Presenter : Kung, Chien-Hao Authors : Medhdi Khashei, Mehdi Bijari 2011, ASOC A novel hybridization of artificial neural.

Documents