This paper works on clustering issues of uncertain time series data prior to prediction process. The aim of uncertainty analysis is to determine how to deal with uncertain data in order to gain knowledge, fit low dimensional model, and to predict. So as to gain a reliable prediction, uncertainty in data could not be ruled out because it may bring important knowledge. Clustering as a step before prediction process can be seen as the most popular representative of unsupervised learning, while classification together with regression are possibly the most frequently considered tasks in supervised learning. Clustering uncertain time series data posts significant challenges on both modeling similarity between uncertain objects and developing efficient computational methods. This work will benefit in many application domains.