The article discusses a method for understanding following relations in time series data. The authors formalize a concept of following motifs between two time series and present a framework to infer these patterns. The proposed framework uses an efficient and scalable method to retrieve motifs from time series, known as the Matrix Profile Method. The framework was tested against several baselines and performed better in simulation datasets. It was able to retrieve following motifs within a pair of time series that two singers sing following each other, and within a pair of time series from two digital currencies, indicating that the values of one currency follow the values of another.
Publication date: 8 Jan 2024
Project Page: https://doi.org/XXXXXXX.XXXXXXX
Paper: https://arxiv.org/pdf/2401.02860
