A Case-Study on the Impact of Dynamic Time Warping in Time Series Regression

TL;DR

This paper investigates the effectiveness of Dynamic Time Warping (DTW) in time series regression, demonstrating its benefits for single-wavelength data and limitations when applied to high-dimensional spectral data.

Contribution

The study provides a detailed case analysis of DTW's impact on spectroscopy time-series regression, highlighting its advantages and limitations in different data scenarios.

Findings

DTW improves regression accuracy for single-wavelength data.

DTW combined with k-NN reveals sample similarities at the time-series level.

Benefits of DTW diminish when using aggregate statistics across many wavelengths.

Abstract

It is well understood that Dynamic Time Warping (DTW) is effective in revealing similarities between time series that do not align perfectly. In this paper, we illustrate this on spectroscopy time-series data. We show that DTW is effective in improving accuracy on a regression task when only a single wavelength is considered. When combined with k-Nearest Neighbour, DTW has the added advantage that it can reveal similarities and differences between samples at the level of the time-series. However, in the problem, we consider here data is available across a spectrum of wavelengths. If aggregate statistics (means, variances) are used across many wavelengths the benefits of DTW are no longer apparent. We present this as another example of a situation where big data trumps sophisticated models in Machine Learning.