Neural Network Modeling For Predicting Rainfall Precipitation

This project aimed at developing a back propagation neural network model to predict rainfall precipitation for Kedah. Rainfall prediction was essential in the Water Management and Control Scheme (WMCS) of Kedah as rainfall precipitation constituted more than 50% of the total water sources to the sta...

Full description

Saved in:
Bibliographic Details
Main Author: Teoh, Boon Wei
Format: Thesis
Language:English
English
Published: 2000
Subjects:
Online Access:https://etd.uum.edu.my/218/1/TEOH_BOON_WEI_-_Neural_network_modeling_for_predicting_rainfall_precipitation.pdf
https://etd.uum.edu.my/218/2/1.TEOH_BOON_WEI_-_Neural_network_modeling_for_predicting_rainfall_precipitation.pdf
https://etd.uum.edu.my/218/
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Universiti Utara Malaysia
Language: English
English
Description
Summary:This project aimed at developing a back propagation neural network model to predict rainfall precipitation for Kedah. Rainfall prediction was essential in the Water Management and Control Scheme (WMCS) of Kedah as rainfall precipitation constituted more than 50% of the total water sources to the state. The back propagation neural network model had been developed using C and Microsoft’s Visual Basics. The data used to train and test the network built was provided by Muda Agricultural Development Authority (MADA). Data obtained consisted of rainfall levels for a maximum of 29 years (1970-l 998) for 31 rainfall stations in Kedah. Upon completion of the training, the best network model produced prediction accuracy of 72.44% for the rainfall levels and this indicated an improvement over the regression approach of 69%. Being the first attempt at predicting the rainfall precipitation in Kedah, the project had succeeded in initiating an application in this area. Further works such as modifying the inputs and the network model could be performed to improve the prediction accuracy of the network.