A hybrid approach model for weather forecasting using multi-task agent

Weather Forecasting is defined and used in the field of knowledge and data engineering simultaneously in order to predict the weather for a specific location. There is large number of various numerical models and algorithm have been developed and enforced to predict the weather forecasting. However some models and algorithms usually do not provide accurate predictions Even though artificial neural networks like unsupervised learning and supervised learning have been considerably applied for predicating the weather forecasting. When considering the multiple neural networks, the redundancy reduction is achieved. In this paper we propose a new hybrid model for weather forecasting, which is based on combination of supervised and unsupervised learning. We address the redundancy issue here and it is overcome by combining these two learning technique with the help of an agent. We also, introduce the accurate prediction of the weather forecasting in this hybrid model. The results presented at the end of paper shows an accurate predication out performance of the proposed method compared to the similar methods in the literature.