The ANN based models are most imperative and demanding for climate
prediction. Back-Propagation Neural Network is an appropriate methodology in
identification of parameters for long-term rainfall data. Presented BPN model
is best to identify climatic parameter for prediction as well as forecast
rainfall. Forecasting of rainfall over vindhya region has been analyzed through
developed BPN model. The observed long period average (LPA) is 956.74 for 45
years data time series of vindhya region. It is observed that the MAD (6.09) is
less than the SD (14.37) during the testing period. Correlation coefficient of
training period is 0.89 and for testing period is 0.95. The performance of the
model is accurately observed. - See more at:
No comments:
Post a Comment