Evaluating DTW Measures via a Synthesis Framework for Time-Series Data

Kishansingh Rajput,Duong Binh Nguyen,Guoning Chen

2024-02-14

Abstract:Time-series data originate from various applications that describe specific observations or quantities of interest over time. Their analysis often involves the comparison across different time-series data sequences, which in turn requires the alignment of these sequences. Dynamic Time Warping (DTW) is the standard approach to achieve an optimal alignment between two temporal signals. Different variations of DTW have been proposed to address various needs for signal alignment or classifications. However, a comprehensive evaluation of their performance in these time-series data processing tasks is lacking. Most DTW measures perform well on certain types of time-series data without a clear explanation of the reason. To address that, we propose a synthesis framework to model the variation between two time-series data sequences for comparison. Our synthesis framework can produce a realistic initial signal and deform it with controllable variations that mimic real-world scenarios. With this synthesis framework, we produce a large number of time-series sequence pairs with different but known variations, which are used to assess the performance of a number of well-known DTW measures for the tasks of alignment and classification. We report their performance on different variations and suggest the proper DTW measure to use based on the type of variations between two time-series sequences. This is the first time such a guideline is presented for selecting a proper DTW measure. To validate our conclusion, we apply our findings to real-world applications, i.e., the detection of the formation top for the oil and gas industry and the pattern search in streamlines for flow visualization.

Machine Learning

What problem does this paper attempt to address?

This paper mainly focuses on the evaluation of Dynamic Time Warping (DTW) methods in time series data analysis. DTW is a method used to find the best alignment between different time series, widely applied in time series matching and analysis. However, there is a lack of comprehensive evaluation of the performance of various DTW variants, especially their performance on specific tasks and data types.

To address this issue, the paper proposes a synthetic framework that can generate pairs of time series with controllable variations, simulating real-world scenarios. This framework consists of two steps: firstly, generating initial signals as realistic as possible, and then generating the second time series with known differences through phase shifting/scaling and random spike insertion/removal.

The paper uses this framework to generate a series of time series pairs with various known variations to evaluate the performance of multiple well-known DTW methods in alignment and classification tasks. They introduce two evaluation metrics, namely Amplitude Distance Measure (ADM) and Time Distance Measure (ADT), to measure the alignment quality. These evaluation results form the guidelines for selecting appropriate DTW methods.

To validate the effectiveness of these guidelines, the researchers apply them to practical applications such as top detection of reservoirs in the oil and gas industry, and pattern searching in flow visualization.

The main contributions of this paper include:
1. A user-controllable synthetic framework that can generate realistic signals and pairs of signals with controllable variations.
2. Comprehensive evaluation of several representative DTW methods on time series pairs with known variations, along with guidelines for selecting DTW methods.

By applying to real data, this paper bridges the gap between DTW variant research and the practical application needs in the selection of methods.