Open AccessSelection of forecast model for consumption (four sectors) and transmission (two Piplines) of natural gas in Punjab (Pakistan) based on ARIMA model
Author(s) -
Maliha Butt
Publication year - 2015
Publication title -
international journal of advanced statistics and probability
Language(s) - English
Resource type - Journals
ISSN - 2307-9045
DOI - 10.14419/ijasp.v3i1.4635
Subject(s) - akaike information criterion , autoregressive integrated moving average , model selection , statistics , correlogram , econometrics , mathematics , mean squared error , box–jenkins , bayesian information criterion , information criteria , selection (genetic algorithm) , standard deviation , time series , computer science , artificial intelligence
The main purpose of this study is to select an appropriate forecast model for Natural Gas Consumption and Transmission System. For ARIMA model, Box-Jenkins Approach (1976) has been adopted i.e. Stationarity of the series has been checked for each data set, correlogram has been estimated for identification of order of ARIMA model and a class of models has been estimated. Then, most adequate and appropriate model is selected by analyzing diagnostics checks. Later on, by comparing values of Akaike Information Criterion (AIC), Schwarz Information Criterion (SIC), and Standard Error (S.E.) of Regression, Root Mean Square Error (RMSC) and Theil Inequality Coefficient (TIC) for each model, forecast model has been finalized. In the end, forecasts have been made using models and compared these forecast values with the actual values for 2010 in order to check the accuracy of the model.