Classificação de Séries Temporais Através de Grafos de Transição de Padrões Ordinais

Regarding its interdisciplinary and broad scope of real-world applications, it is evident the need of extracting knowledge from time series data. Mining this type of data, however, faces several complexities due to its unique properties. In this work, we propose a new feature, retained from the Ordinal Pattern Transition Graph, called probability of self-transition. Our proposal was tested in a real Urban Computing problem, referred to the classification of transportation mode. We had better accuracy results than well-known Information Theory features, besides being less dependent on Ordinal Patterns parameters.