论文信息 - On The Rainfall Time Series Prediction Using Multilayer Perceptron Artificial Neural Network

On The Rainfall Time Series Prediction Using Multilayer Perceptron Artificial Neural Network

148  Abstract— In this paper, Multilayer Perceptron Neural Network is proposed as an intelligent tool for predicting Rainfall Time Series. A Rainfall samples have been collected from the authorized Government Rainfall monitoring agency in Yavatmal, Maharashtra state, India. Rainfall percentage rise and fall with an irregular cycle.Multi-step ahead (1, 5, 10, 20) predictions of this Rainfall Data series have been carried out using the proposed Multilayer Perceptron Neural Network. It is seen that the performance measures such as MSE (Mean square error), and NMSE (Normalized mean square error) on testing as well as training data set for short term prediction are found as optimal in comparison with other network such as Jordon Elmann Neural Network, SOFM (Self organized feature map), RNN (Recurrent neural network).Out of all these networks, the results for Jordon Elman and MLP network were by far the closest. Hence, in this paper the analysis of these two networks is shown. The different results Parameters are calculated by using software, “Neurosolution 5.0”. Neurosolution is an object oriented environment for designing, prototyping, simulating, and deploying artificial neural network (ANN) solutions. It is also evident that estimated neural network model closely follows the actual outputs for desired outputs for multi step ahead predictions [1].

Rohit R. Deshpande

[1] J J Hopfield,et al. Neural networks and physical systems with emergent collective computational abilities. , 1982, Proceedings of the National Academy of Sciences of the United States of America.

[2] Rohit R. Deshpande,et al. ON THE SUNSPOT TIME SERIES PREDICTION USING JORDON ELMAN ARTIFICIAL NEURAL NETWORK (ANN) , 2011 .

[3] Teuvo Kohonen,et al. The self-organizing map , 1990, Neurocomputing.

[4] Geoffrey E. Hinton,et al. A time-delay neural network architecture for isolated word recognition , 1990, Neural Networks.

[5] Michael I. Jordan. Attractor dynamics and parallelism in a connectionist sequential machine , 1990 .

[6] S. M. Gulhane,et al. Multi-Step Ahead Prediction Of Chaotic Time Series Using Recurrent Neural Network Models , 2011 .

[7] R. Lippmann,et al. An introduction to computing with neural nets , 1987, IEEE ASSP Magazine.

[8] Robert A. Jacobs,et al. Increased rates of convergence through learning rate adaptation , 1987, Neural Networks.

[9] J. Makhoul. Pattern recognition properties of neural networks , 1991, Neural Networks for Signal Processing Proceedings of the 1991 IEEE Workshop.