Scalable End-to-end Recurrent Neural Network for Variable star classification

@article{Becker2020ScalableER,
  title={Scalable End-to-end Recurrent Neural Network for Variable star classification},
  author={Ignacio Becker and Karim Pichara and M{\'a}rcio Catel{\'a}n and Pavlos Protopapas and Carlos Aguirre and Fatemeh Nikzat},
  journal={ArXiv},
  year={2020},
  volume={abs/2002.00994}
}
During the last decade, considerable effort has been made to perform automatic classification of variable stars using machine-learning techniques. Traditionally, light curves are represented as a vector of descriptors or features used as input for many algorithms. Some features are computationally expensive, cannot be updated quickly and hence for large data sets such as the LSST cannot be applied. Previous work has been done to develop alternative unsupervised feature extraction algorithms… 
Classification of Variable Stars Light Curves Using Long Short Term Memory Network
TLDR
This work presents a new automated method, 1D CNN-LSTM, for classifying variable stars using a hybrid neural network of one-dimensional CNN and LSTM network which employs the raw time-series data from the variable stars.
Light-curve classification with recurrent neural networks for GOTO: dealing with imbalanced data
TLDR
A recurrent neural network (RNN) classifier that takes in photometric time-series data and additional contextual information to produce real-time classification of objects observed by the Gravitational-wave Optical Transient Observer (GOTO), and use an algorithm-level approach for handling imbalance with a focal loss function.
Image-based Classification of Variable Stars: First Results from Optical Gravitational Lensing Experiment Data
TLDR
This work uses an image-based Convolutional Neural Network to classify the different types of variable stars, using images of phase-folded light curves from the OGLE-III survey for training, validating and testing and used OGL survey as an independent data set for testing.
The effect of phased recurrent units in the classification of multiple catalogs of astronomical lightcurves
TLDR
This work uses seven catalogs containing periodic and nonperiodic astronomical objects and shows that LSTM outperformed PLSTM on 6/7 datasets and the combination of both units enhances the results in all datasets.
Classification of Periodic Variable Stars with Novel Cyclic-Permutation Invariant Neural Networks
TLDR
Cyclic-permutation invariant networks are presented, a novel class of NNs for which invariance to phase shifts is guaranteed through polar coordinate convolutions, which are implemented by means of "Symmetry Padding".
Cyclic-Permutation Invariant Networks for Modeling Periodic Sequential Data: Application to Variable Star Classification
TLDR
Across three different datasets of periodic variable star light-curves, it is shown that two particular implementations of the cyclic-permutation invariant network consistently outperform both non-invariant versions of the same network, as well as previous state-of-the-art approaches including Random Forests.
A survey on machine learning based light curve analysis for variable astronomical sources
TLDR
This survey reviews important developments in light Curve analysis over the past years, summarizes the basic concepts in machine learning and their applications in light curve analysis and concludes perspectives and challenges for light curveAnalysis in the near future.
Informative Bayesian model selection for RR Lyrae star classifiers
TLDR
This paper develops a method based on an informative marginal likelihood to evaluate variable star classifiers that outperforms traditional non-informative cross-validation strategies, even when penalized models are assessed.
Classification of Fermi-LAT sources with deep learning using energy and time spectra
TLDR
A new deep neural network approach for the classification of unidentified or unassociated gamma-ray sources in the last release of the Fermi-LAT catalogue (4FGL-DR2) obtained with 10 yr of data, which provides powerful classifiers, with a performance that is comparable to previous analyses based on human-crafted features.
Development of accurate classification of heavenly bodies using novel machine learning techniques
TLDR
A novel methodology that enables an efficient and accurate classification of cosmic objects (3 classes) based on evolutionary optimization of classifiers based on genetic algorithm implemented on the data which helps to find the optimal parameters for the classifiers.
...
...

References

SHOWING 1-10 OF 61 REFERENCES
Deep multi-survey classification of variable stars
TLDR
A novel Deep Learning model based on convolutional units that captures the essential classification patterns regardless of cadence and filter in light curve classification and introduces a novel data augmentation schema for unevenly sampled time series.
Clustering Based Feature Learning on Variable Stars
TLDR
This work presents what is, to their knowledge, the first unsupervised feature learning algorithm designed for variable stars, which first extracts a large number of lightcurves from a given set of photometric data, which are then clustered to find common local patterns in the time series.
A recurrent neural network for classification of unevenly sampled variable stars
TLDR
A novel unsupervised autoencoding recurrent neural network that makes explicit use of sampling times and known heteroskedastic noise properties and produces state-of-the-art supervised classification models that rival other best-in-class approaches.
Deep Recurrent Neural Networks for Supernovae Classification
TLDR
Using the Supernovae Photometric Classification Challenge (SPCC) data, it is found that deep networks are capable of learning about light curves, however the performance of the network is highly sensitive to the amount of training data.
ON MACHINE-LEARNED CLASSIFICATION OF VARIABLE STARS WITH SPARSE AND NOISY TIME-SERIES DATA
TLDR
A methodology for variable-star classification, drawing from modern machine-learning techniques, which is effective for identifying samples of specific science classes and presents the first astronomical use of hierarchical classification methods to incorporate a known class taxonomy in the classifier.
Automatic Survey-invariant Classification of Variable Stars
TLDR
A full Probabilistic model is proposed that represents the joint distribution of features from two surveys as well as a probabilistic transformation of the features between one survey to the other and to effectively run a variable star classification model in a new survey.
Unsupervised classification of variable stars
TLDR
This work proposes a query based method where astronomers can find groups of variable stars ranked by similarity, and develops a fast similarity function specific for light curves, based on a novel data structure that allows scaling the search over the entire dataset of unlabelled objects.
Rotation-invariant convolutional neural networks for galaxy morphology prediction
TLDR
A deep neural network model for galaxy morphology classification which exploits translational and rotational symmetry is developed in the context of the Galaxy Challenge, an international competition to build the best model for morphology classification based on annotated images from the Galaxy Zoo project.
Convolutional neural networks for transient candidate vetting in large-scale surveys
TLDR
This work presents an approach based on convolutional neural networks to discriminate between true astrophysical sources and artefacts in reference-subtracted optical images and shows that relatively simple networks are already competitive with state-of-the-art systems and that their quality can be improved via slightly deeper networks and additional preprocessing steps.
Meta-classification for Variable Stars
TLDR
A new meta-model that automatically integrates existing classification models of variable stars is proposed that incorporates existing models that are trained in a different context, answering different questions and using different representations of data.
...
...