Pattern Discovery in Time Series with Byte Pair Encoding
This addresses the need for interpretable time series analysis in health monitoring, though it is incremental as it adapts an existing compression technique to a new domain.
The authors tackled the problem of analyzing noisy, variable-length time series from wearable sensors by proposing an unsupervised method based on Byte Pair Encoding to identify interpretable patterns, and demonstrated that it outperforms state-of-the-art approaches on a real-world dataset.
The growing popularity of wearable sensors has generated large quantities of temporal physiological and activity data. Ability to analyze this data offers new opportunities for real-time health monitoring and forecasting. However, temporal physiological data presents many analytic challenges: the data is noisy, contains many missing values, and each series has a different length. Most methods proposed for time series analysis and classification do not handle datasets with these characteristics nor do they offer interpretability and explainability, a critical requirement in the health domain. We propose an unsupervised method for learning representations of time series based on common patterns identified within them. The patterns are, interpretable, variable in length, and extracted using Byte Pair Encoding compression technique. In this way the method can capture both long-term and short-term dependencies present in the data. We show that this method applies to both univariate and multivariate time series and beats state-of-the-art approaches on a real world dataset collected from wearable sensors.