Rainfall forecasting by relevant attributes using artificial neural networks - a comparative study
by Saroj Kr. Biswas; Leniency Marbaniang; Biswajit Purkayastha; Manomita Chakraborty; Heisnam Rohen Singh; Monali Bordoloi
International Journal of Big Data Intelligence (IJBDI), Vol. 3, No. 2, 2016

Abstract: Many models have been developed for rainfall forecasting from time to time. Artificial neural networks (ANNs) using back propagation algorithm, are the most popular and widely used forecasting models in the present decade. Rainfall depends on many weather parameters (attributes) including pressure, temperature, wind speed, and so on. A set of attributes is habitually used for rainfall prediction, which consists of relevant and irrelevant attributes and from the viewpoint of managing a dataset which can be enormous. Hence, reducing the number of attributes by selecting only the relevant ones is desirable. Doing so, higher performance with lower computational effort is expected. Therefore, feature selection needs to be done to identify the effective parameters to improve the forecasting ability of a model. Consequently, this paper proposes a feature selection algorithm for rainfall forecasting using neural network and investigates the performance of different ANN methods such as multi-layer feed forward neural network (MLFNN), radial basis function neural network (RBFNN), focused time delay neural network (FTDNN) and nonlinear autoregressive exogenous input neural network (NARXNN). From the empirical results, it is observed that NARXNN produces better predicted accuracy than others. It is also observed that the proposed model outperforms one earlier forecasting model.

Online publication date: Wed, 29-Jun-2016

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Big Data Intelligence (IJBDI):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com