Humans always want to know what will happen in the next days or future. Now a days, prediction is most important that gives a knowledge about what will happen in the future. Today, weather is variable in nature so its prediction has become an important field of research. Several traditional methods are used for weather prediction. But they have their limitations for handling and predicting accurately. Predicting weather using artificial neural network (ANN) gives better results than the traditional methods. Stochastic weight updation with reinforced learning is used for the learning of neural network. There are several advantages by using this method, its implementation is simple for network topology and it allow better parallelization of the backpropagation algorithm. In stochastic weight update method, weights are selected according to certain threshold for updation instead of updating all weights. Reinforced learning is used for balanced weight updation. Here for attaining an optimized and accurate result it is used along with the stochastic weight updation.
Published In:IJCSN Journal Volume 8, Issue 3
Date of Publication : June 2019
Pages : 217-221
Tables : 01
Siji Chacko :
received her B.Tech (CSE) degree from University of Kerala in 2017. She is currently pursuing her Masters in Computer Science & Engineering from KTU.
Sam G Benjamin :
is working as Assistant Professor in computer science and engineering Department. His research interests focuses on image processing, data mining, and image mining. He has published several papers on image processing.
An effective method for the weather prediction is discussed in this paper. The main advantage of this method is the low computational complexity, cheap implementation cost and accurate result. Now, weather prediction has a big challenge of predicting the accurate results. The difficulty of this depends on the complex nature of parameters, variable weather and absence of seasonal change. By using ANN these problems can be avoided. It accepts all complex parameters and generates the intelligent patterns during training and it uses the same patterns to generate the forecasts. Stochastic weight update with reinforced learning shows greater accuracy in prediction and it can be used as a powerful tool for weather prediction.