Decision making for uncertain data in dynamic environment using hybrid method

Predicting plays a main role in making a good decision in a dynamic environment with real time data, such as weather predictions that normally are heterogeneous, various and vast. The use of heterogeneous of data available causes increasing uncertainty. Vast amount of data from a dynamic environment will be very uncertain and the prediction of future behavior of such a situation is very hard and if it is possible, the result will not be accurate and may lead to incorrect predictions, Therefore, for the problem statement we propose a new framework that is called hybrid method based on Fuzzy c-mean clustering and Type-2 fuzzy logic system with gradient descent algorithm for training data that would be applicable in a dynamic environment, the hybrid method is able to manage uncertain data such as outliers and noise to get more accurate results and applicable in dynamic environment as well. The results proved the hybrid method is usable and more accurate than the based method (Type-2 FLS).