Semi-supervised Learning for Photometric Supernova Classification

Abstract

We present a semi-supervised method for photometric supernova typing. Our approach is to first use the nonlinear dimension reduction technique diffusion map to detect structure in a database of supernova light curves and subsequently employ random forest classification on a spectroscopically confirmed training set to learn a model that can predict the type of each newly observed supernova. We demonstrate that this is an effective method for supernova typing. As supernova numbers increase, our semi-supervised method efficiently utilizes this information to improve classification, a property not enjoyed by template based methods. Applied to supernova data simulated by Kessler et al. (2010b) to mimic those of the Dark Energy Survey, our methods achieve (cross-validated) 95% Type Ia purity and 87% Type Ia efficiency on the spectroscopic sample, but only 50% Type Ia purity and 50% efficiency on the photometric sample due to their spectroscopic follow-up strategy. To improve the performance on the photometric sample, we search for better spectroscopic follow-up procedures by studying the sensitivity of our machine learned supernova classification on the specific strategy used to obtain training sets. With a fixed amount of spectroscopic follow-up time, we find that deeper magnitude-limited spectroscopic surveys are better for producing training sets. For supernova Ia (II-P) typing, we obtain a 44% (1%) increase in purity to 72% (87%) and 30% (162%) increase in efficiency to 65% (84%) of the sample using a 25th (24.5th) magnitude-limited survey instead of the shallower spectroscopic sample used in the original simulations. When redshift information is available, we incorporate it into our analysis using a novel method of altering the diffusion map representation of the supernovae. Incorporating host redshifts leads to a 5% improvement in Type Ia purity and 13% improvement in Type Ia efficiency.

16 pages, 11 figures, accepted for publication in MNRAS

References in corpus (8)

• On Machine-Learned Classification of Variable Stars with Sparse and Noisy Time-Series Data
• Random forest automated supervised classification of Hipparcos periodic variable stars
• Results from the Supernova Photometric Classification Challenge
• Bayesian Estimation Applied to Multiple Species: Towards cosmology with a million supernovae
• Bayesian Single-Epoch Photometric Classification of Supernovae
• Type II-P Supernovae as Standard Candles: The SDSS-II Sample Revisited
• Statistical Classification Techniques for Photometric Supernova Typing
• Characterizing the contaminating distance distribution for Bayesian supernova cosmology

Cited by in corpus (37)

• Photometric Supernova Classification With Machine Learning
• RAPID: Early Classification of Explosive Transients using Deep Learning
• The weirdest SDSS galaxies: results from an outlier detection algorithm
• Using Machine Learning for Discovery in Synoptic Survey Imaging
• Cosmology with Photometrically-Classified Type Ia Supernovae from the SDSS-II Supernova Survey
• Avocado: Photometric Classification of Astronomical Transients with Gaussian Process Augmentation
• SuperRAENN: A Semi-supervised Supernova Photometric Classification Pipeline Trained on Pan-STARRS1 Medium Deep Survey Supernovae
• Computational Intelligence Challenges and Applications on Large-Scale Astronomical Time Series Databases
• Optimizing spectroscopic follow-up strategies for supernova photometric classification with active learning
• Catalogues of Active Galactic Nuclei From Gaia and unWISE Data
• Kernel PCA for type Ia supernovae photometric classification
• Active Learning to Overcome Sample Selection Bias: Application to Photometric Variable Star Classification
• Supernova Photometric Classification Pipelines Trained on Spectroscopically Classified Supernovae from the Pan-STARRS1 Medium-Deep Survey
• Star-Galaxy Classification in Multi-Band Optical Imaging
• A simple and robust method for automated photometric classification of supernovae using neural networks
• PELICAN: deeP architecturE for the LIght Curve ANalysis
• Probing the stellar initial mass function with high-$z$ supernovae
• A review of unsupervised learning in astronomy
• Evidence for Dust Destruction from the Early-time Colour Change of GRB 120119A
• STACCATO: A Novel Solution to Supernova Photometric Classification with Biased Training Sets
• A probabilistic approach to emission-line galaxy classification
• Exploring the spectroscopic diversity of type Ia supernovae with DRACULA: a machine learning approach
• Defining Photometric Peculiar Type Ia Supernovae
• Fink: early supernovae Ia classification using active learning
• Photometric Classification of 2315 Pan-STARRS1 Supernovae with Superphot
• The Twins Embedding of Type Ia Supernovae I: The Diversity of Spectra at Maximum Light
• Photometric Classification of Early-Time Supernova Lightcurves with SCONE
• deepSIP: Linking Type Ia Supernova Spectra to Photometric Quantities with Deep Learning
• Parameter Estimation with BEAMS in the presence of biases and correlations
• The Dark Knight Falters
• Data challenges as a tool for time-domain astronomy
• Statistical Study of 2XMMi-DR3/SDSS-DR8 Cross-correlation Sample
• Stratified Learning: A General-Purpose Statistical Method for Improved Learning under Covariate Shift
• An Algorithm for the Visualization of Relevant Patterns in Astronomical Light Curves
• Real-time Light Curve Classification Framework for the Wide Field Survey Telescope Using Modified Semi-supervised Variational Auto-Encoder
• The snapshot distance method: estimating the distance to a Type Ia supernova from minimal observations
• Radio Galaxy Zoo: Using semi-supervised learning to leverage large unlabelled data-sets for radio galaxy classification under data-set shift