The generation of qualitative descriptions of multivariate time series using fuzzy logic

• An approach to generating qualitative descriptions of multivariate time series using the “Temporal Fuzzy Model” as an intermediate representation is presented.
• We present a formal description of the temporal fuzzy models, the induction and inference algorithms associated with them, and a process that acquires text from them.
• We present a set of experiments in order to establish a measure of how valid a representation is the model of the multivariate system, and also a second set of experiments that validate the qualitative description process.

An approach to generating qualitative descriptions of multivariate time series using the ‘Temporal Fuzzy Model’ as an intermediate representation is presented. This model, based on fuzzy logic, allows of representing in a linguistic way, the information contained in multivariate time series. The extraction of qualitative descriptions can then be carried out directly. We present a formal description of the temporal fuzzy models, the induction and inference algorithms associated with them, and a process that acquires text from them. Finally, we present a set of experiments in order to establish a measure of how valid a representation is the model of the multivariate system, and also a second set of experiments that validate the qualitative description process.

Figure optionsDownload as PowerPoint slide