Pattern analysis approach for prediction using Wavelet Neural Networks

Temperature Prediction has gained wide application and growing interest. It has gained importance in the studies by various researchers in this field. In this paper, we study the chaotic behavior of Temperature Prediction using Time Series Analysis and Pattern Mining. The proposed method of Wavelet Neural Network makes use of the temperature pattern observed in the past during a part of the year to predict the temperature observed at the same part in the following year with high degree of accuracy. This paper investigates the temperature data collected in Taipei for the year 1995 and 1996 and examines via comparing with the existing techniques whether the proposed method can provide the required level of performance, which is sufficiently good and robust so as to provide a reliable forecast model. Results obtained are demonstrated to validate the generalization ability and efficiency of the proposed model over the simulated network. Experimental results have shown that the model considered could represent the Temperature Prediction behavior very accurately. Thus outperforming existing approaches.