Improved Photometric Classification of Supernovae using Deep Learning

Moss, Adam

doi:10.48550/arxiv.1810.06441

Cited by 14 publications

(22 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Extracted features can be used to classify supernovae using sequential cuts (Bazin et al 2011;Campbell et al 2013) or machine learning algorithms (Kessler et al 2010b;Möller et al 2016;Lochner et al 2016;Dai et al 2018). Recent advances in Deep Learning, a branch of machine learning, have shown that neural network classifiers trained on raw data can outperform classifiers based on handcrafted features (Charnock & Moss 2017;Kimura et al 2017;Moss 2018). In this work, we explore this promising direction and obtain state-of-the-art results on a variety of supernovae classification tasks.…”

Section: Introductionmentioning

confidence: 98%

SuperNNova: an open-source framework for Bayesian, neural network-based supernova classification

Möller

Boissière²

2019

Monthly Notices of the Royal Astronomical Society

135

116

View full text Add to dashboard Cite

We introduce SuperNNova, an open source supernova photometric classification framework which leverages recent advances in deep neural networks. Our core algorithm is a recurrent neural network (RNN) that is trained to classify light-curves using photometric information only. Additional information such as host-galaxy redshift can be incorporated to improve performance. We evaluate our framework using realistic supernovae simulations that include survey detection. We show that our method, for the type Ia vs. non Ia supernovae classification problem, reaches accuracies greater than 96.92 ± 0.09 without any redshift information and up to 99.55 ± 0.06 when redshift, either photometric or spectroscopic, is available. Further, we show that our method attains unprecedented performance for classification of incomplete light-curves, reaching accuracies > 86.4 ± 0.1 (> 93.5 ± 0.8) without host-galaxy redshift (with redshift information) two days before maximum light. In contrast with previous methods, there is no need for time-consuming feature engineering and we show that our method scales to very large datasets with a modest computing budget. In addition, we investigate often neglected pitfalls of machine learning algorithms. We show that commonly used algorithms suffer from poor calibration and overconfidence on out-of-distribution samples when applied to supernovae data. We devise extensive tests to estimate the robustness of classifiers and cast the learning procedure under a Bayesian light, demonstrating a much better handling of uncertainties. We study the benefits of Bayesian RNNs for SN Ia cosmology. Our code is open-sourced and available in github ‡.

show abstract

Section: Introductionmentioning

confidence: 98%

SuperNNova: an open-source framework for Bayesian, neural network-based supernova classification

Möller

Boissière²

2019

Monthly Notices of the Royal Astronomical Society

135

116

View full text Add to dashboard Cite

show abstract

“…Deep learning is a branch of machine learning that seeks to eliminate the necessity of human-designed features, decreasing the computational cost as well as avoiding the introduction of potential biases (Charnock & Moss 2017;Moss 2018;Naul et al 2018;Aguirre et al 2019). In recent years, many deep learning techniques have been applied to the challenge of photometric SN classification.…”

Section: Photometric Supernova Classificationmentioning

confidence: 99%

Photometric Classification of Early-Time Supernova Lightcurves with SCONE

Qu,

Sako

2021

Preprint

View full text Add to dashboard Cite

In this work, we present classification results on early supernova lightcurves from SCONE, a photometric classifier that uses convolutional neural networks to categorize supernovae (SNe) by type using lightcurve data. SCONE is able to identify SN types from lightcurves at any stage, from the night of initial alert to the end of their lifetimes. Simulated LSST SNe lightcurves were truncated at 0, 5, 15, 25, and 50 days after the trigger date and used to train Gaussian processes in wavelength and time space to produce wavelength-time heatmaps. SCONE uses these heatmaps to perform 6-way classification between SN types Ia, II, Ibc, Ia-91bg, Iax, and SLSN-I. SCONE is able to perform classification with or without redshift, but we show that incorporating redshift information improves performance at each epoch. SCONE achieved 75% overall accuracy at the date of trigger (60% without redshift), and 89% accuracy 50 days after trigger (82% without redshift). SCONE was also tested on bright subsets of SNe (r < 20 mag) and produced 91% accuracy at the date of trigger (83% without redshift) and 95% 5 days after trigger (94.7% without redshift). SCONE is the first application of convolutional neural networks to the early-time photometric transient classification problem. All of the data processing and model code developed for this paper can be found in the SCONE software package located at github.com/helenqu/scone (Qu 2021).

show abstract

“…Most approaches involve lightcurve template matching (Sako et al 2011), or feature extraction paired with either sequential cuts (Bazin et al 2011;Campbell et al 2013) or machine learning algorithms (Möller et al 2016;Lochner et al 2016;Dai et al 2018;Boone 2019). Most recently, the spotlight has been on deep learning techniques since it has been shown that classification based on handcrafted features is not only more time-intensive for the researcher but is outperformed by neural networks trained on raw data (Charnock & Moss 2017;Moss 2018;Kimura et al 2017). Since then, many neural network architectures have been explored for SN photometric classification, such as PELICAN's CNN architecture (Pasquet et al 2019) and SuperNNova's deep recurrent network (Möller & de Boissière 2020).…”

Section: Introductionmentioning

confidence: 99%

SCONE: Supernova Classification with a Convolutional Neural Network

Qu,

Sako,

Möller

et al. 2021

Preprint

View full text Add to dashboard Cite

We present a novel method of classifying Type Ia supernovae using convolutional neural networks, a neural network framework typically used for image recognition. Our model is trained on photometric information only, eliminating the need for accurate redshift data. Photometric data is pre-processed via 2D Gaussian process regression into two-dimensional images created from flux values at each location in wavelength-time space. These "flux heatmaps" of each supernova detection, along with "uncertainty heatmaps" of the Gaussian process uncertainty, constitute the dataset for our model. This preprocessing step not only smooths over irregular sampling rates between filters but also allows SCONE to be independent of the filter set on which it was trained. Our model has achieved impressive performance without redshift on the in-distribution SNIa classification problem: 99.73 ± 0.26% test accuracy with no over/underfitting on a subset of supernovae from PLAsTiCC's unblinded test dataset. We have also achieved 98.18±0.3% test accuracy performing 6-way classification of supernovae by type. The out-of-distribution performance does not fully match the in-distribution results, suggesting that the detailed characteristics of the training sample in comparison to the test sample have a big impact on the performance. We discuss the implication and directions for future work. All of the data processing and model code developed for this paper can be found in the SCONE software package located at github.com/helenqu/scone (Qu 2021).

show abstract

Improved Photometric Classification of Supernovae using Deep Learning

Cited by 14 publications

References 15 publications

SuperNNova: an open-source framework for Bayesian, neural network-based supernova classification

SuperNNova: an open-source framework for Bayesian, neural network-based supernova classification

Photometric Classification of Early-Time Supernova Lightcurves with SCONE

SCONE: Supernova Classification with a Convolutional Neural Network

Contact Info

Product

Resources

About