An effective fuzzy C-mean and type-2 fuzzy logic for weather forecasting.

Meteorological forecasting is applicable for versatile applications. Accurate weather prediction saves lives, money and time in both local and global area. Forecasting accuracy is still not accurate because of the uncertain (fuzzy) data of nature, due to several reasons including: incomplete data, hand writing error, device error, precision of measurements and discreet description of connective phenomena Inherent part reflecting our understanding of things. On the other hand in global area with large amount of data to process whole the data is time consuming, thus, to improve the quality of data and execution time, we need to manage the uncertainty of data and extract desired data. Therefore the uncertainty management and process the data demand intelligent methods with knowledge based approaches. This paper reviews challenges in this field and compares advantages and drawbacks of the existing methods that essentially are only applicable for local area. Finally we proposed a hybrid technique for new research based on fuzzy c-mean clustering technique and type-2 fuzzy logic that is useable in both local and global area. Finally we show our experiments and prove that hybrid technique performs better than existing weather prediction methods in low error rate.

Actions (login required)

View Item