Light curve classification using distance metrics

The rise of synoptic sky surveys has ushered in an era of big data in time-domain astronomy, making data science and machine learning essential tools for studying celestial objects. Tree-based (e.g., Random Forests) and deep-learning models represent the current standard in the field. The direct use of distance metrics is an approach that has not been explored in time-domain astronomy, but distance-based methods can aid in increasing the interpretability of the classification result and decrease the computational costs. In this paper, we present an investigation of distance metrics for the classification of astrophysical time series, or light curves. In particular, we classify light curves of variable stars by comparing the distances between objects of different classes. We develop a distance-based classifier and demonstrate its use for classification and dimensionality reduction exploring the performance of 18 distance metrics applied to a catalog of 700,000 variable stars in 10 classes. We show that this classifier meets state-of-the-art performance but has lower computational requirements and improved interpretability.