Modelling and forecasting the marine fish production of India
250 / 185
Authors
-
G Sunilkumar
Indian Agricultural Statistics Research Institute, New Delhi - 110 012, India
-
PRAJNESHU ,
Indian Agricultural Statistics Research Institute, New Delhi - 110 012, India
Abstract
The newly emerging area of ''Wavelet analysis'' is thoroughly discussed along with its modification through different
thresholding procedures. As an illustration, the methodology is applied to model and forecast country’s marine fish production
data during 1971 to 2002 using S-PLUS Wavelet toolkit. Presence of autocorrelation for residuals is detected by Durbin-
Watson test. Subsequently, Wavelet with data-dependent thresholding as well as other approaches,
viz.
, Box-Jenkins ARIMA
and non-parametric regression techniques, are also applied to the data. Comparison, on the basis of mean square errors,
shows superiority of Wavelet procedure over Box-Jenkins ARIMA and non-parametric regression method for data under
consideration. Finally, forecast of country’s marine fish production for the year 2003 is obtained through Wavelet methodology.
