Fuzzy Relational Representation and Interpretation of Temporal Data

The interpretation and description of time series play a major role in better approximation of multivariable time series, which is central to the improvements of the performance of classifiers and predictors. Granular time series are models of time series formed at the level of information granules that are expressed by the representation of space and time realized in the form of information granules. In the description, information granules are constructed over the space of amplitude and change of amplitude giving rise to a suitable and flexible level of abstraction. Subsequently, different granular logic models (viz. those built at varying levels of granularity) are investigated and compared with respect to their interpretability and accuracy.