This paper develops a new time series clustering procedure allowing for heteroskedasticity, non-normality and model's non-linearity. At this aim, we follow fuzzy approach. Specifically, considering Dynamic Conditional Score (DCS) model, propose to cluster according their estimated conditional moments via the Autocorrelation-based C-means (A-FCM) algorithm. The DCS parametric modeling is appeali...

Four different approaches to robust fuzzy clustering of time series are presented and compared with respect other existent approaches. These useful cluster when outlying values found in these series, which is often the rule most real data applications. A representation by using B-splines considered and, later, methods applied on fitted coefficients. Feasible algorithms for implementing methodol...