PHOTOMETRIC SUPERNOVA CLASSIFICATION WITH MACHINE LEARNING

By Michelle Lochner, Jason D. McEwen, Hiranya V. Peiris, Ofer Lahav, Max K. Winter. 2016

168 citations recorded by the source corpus.

Automated photometric supernova classification has become an active area of research in recent years in light of current and upcoming imaging surveys such as the Dark Energy Survey (DES) and the Large Synoptic Survey Telescope, given that spectroscopic confirmation of type for all supernovae discovered will be impossible. Here, we develop a multi-faceted classification pipeline, combining existing and new approaches. Our pipeline consists of two stages: extracting descriptive features from the light curves and classification using a machine learning algorithm. Our feature extraction methods vary from model-dependent techniques, namely SALT2 fits, to more independent techniques that fit parametric models to curves, to a completely model-independent wavelet approach. We cover a range of representative machine learning algorithms, including naive Bayes, k-nearest neighbors, support vector machines, artificial neural networks, and boosted decision trees (BDTs). We test the pipeline on simulated multi-band DES light curves from the Supernova Photometric Classification Challenge. Using the commonly used area under the curve (AUC) of the Receiver Operating Characteristic as a metric, we find that the SALT2 fits and the wavelet approach, with the BDTs algorithm, each achieve an AUC of 0.98, where 1 represents perfect classification. We find that a representative training set is essential for good classification, whatever the feature set or algorithm, with implications for spectroscopic follow-up. Importantly, we find that by using either the SALT2 or the wavelet feature sets with a BDT algorithm, accurate classification is possible purely from light curve data, without the need for any redshift information.

Subfields: Artificial Intelligence, Astronomy and Astrophysics, Nuclear and High Energy Physics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Francisco Javier Sanchez-Lopez
2. David Kirkby
3. A. Ihler
4. A. Riess
5. Noble Kennamer
6. R. Kirshner
7. E. Berger

In Astronomy and Astrophysics

1. A. Riess
2. Ryan Chornock
3. S. Smartt
4. R. Kirshner
5. A. Filippenko
6. E. Berger
7. J. Tonry
8. R. Foley

In Nuclear and High Energy Physics

1. Ryan Chornock
2. A. Riess
3. R. Kirshner
4. S. Smartt
5. E. Berger
6. D. Scolnic
7. J. Tonry
8. R. Kessler

Citations in the corpus

• Correcting Type Ia Supernova Distances for Selection Biases and Contamination in Photometrically Identified Samples (2016, methodology): “Numerous machine-learning (ML) methods can be applied to photometric classification as described in Lochner et al. (2016).”
• Measuring Dark Energy Properties with Photometrically Classified Pan-STARRS Supernovae. II. Cosmological Parameters (2017, background): “Machine learning algorithms in particular have been found to yield both efficiencies (few bona fide SNe Ia are misclassified) and sample purities &96% in cases where the classifier can be trained on a representative SN sample (Sako et al. 2014; Lochner et al. 2016).”
• ContextNet: Deep learning for Star Galaxy Classification (2018, background): “…and applications, including a deep learning approach for analyzing strongly lensed systems (Hezaveh et al., 2017), a probabilistic graphical model for processing astronomical images (Regier et al., 2015), an ensemble approach for classification of supernova (Lochner et al., 2016), and many others.”