Abstract:In recent years the amount of publicly available astronomical data has increased exponentially, with a remarkable example being large scale multiepoch photometric surveys. This wealth of data poses challenges to the classical methodologies commonly employed in the study of variable objects. As a response, deep learning techniques are increasingly being explored to effectively classify, analyze, and interpret these large datasets. In this paper we use two-dimensional histograms to represent Optical Gravitational Lensing Experiment (OGLE) phasefolded light curves as images. We use a Convolutional Neural Network (CNN) to classify variable objects within eight different categories (from now on labels): Classical Cepheid (CEP), RR Lyrae (RR), Long Period Variable (LPV), Miras (M), Ellipsoidal Binary (ELL), Delta Scuti (DST), Eclipsing Binary (E), and spurious class with Incorrect Periods (Rndm). We set up different training sets to train the same CNN architecture in order to characterize the impact of the training. The training sets were built from the same source of labels but different filters and balancing techniques were applied. Namely: Undersampling (U), Data Augmentation (DA), and Batch Balancing (BB). The best performance was achieved with the BB approach and a training sample size of $\sim$370000 stars. Regarding computational performance, the image representation production rate is of $\sim$76 images per core per second, and the time to predict is $\sim$ 60$\, \mu\text{s}$ per star. The accuracy of the classification improves from $\sim$ 92%, when based only on the CNN, to $\sim$ 98% when the results of the CNN are combined with the period and amplitude features in a two step approach. This methodology achieves comparable results with previous studies but with two main advantages: the identification of miscalculated periods and the improvement in computational time cost.

Deep learning of quasar lightcurves in the LSST era

Predicting High Magnification Events in Microlensed Quasars in the Era of LSST using Recurrent Neural Networks

Joint Modeling of Quasar Variability and Accretion Disk Reprocessing using Latent Stochastic Differential Equations

Latent Stochastic Differential Equations for Modeling Quasar Variability and Inferring Black Hole Properties

The LSST AGN Data Challenge: Selection methods

Detecting unresolved lensed SNe Ia in LSST using blended light curves

Conditional Neural Process for Nonparametric Modeling of Active Galactic Nuclei Light Curves

Deep learning for studies of galaxy morphology

The PAU survey: Estimating galaxy photometry with deep learning

Discovering Ca II Absorption Lines With a Neural Network

Resolving the Vicinity of Supermassive Black Holes with Gravitational Microlensing

Using Convolutional Neural Networks to Search for Strongly Lensed Quasars in KiDS DR5

A deep learning approach to quasar continuum prediction

Deep learning from 21-cm tomography of the Cosmic Dawn and Reionization

Application of Convolutional Neural Networks to time domain astrophysics. 2D image analysis of OGLE light curves

Deep Learning Voigt Profiles. I. Single-Cloud Doublets

Lens Modeling of STRIDES Strongly Lensed Quasars using Neural Posterior Estimation

Deep Learning of Quasar Spectra to Discover and Characterize Damped Lyα Systems

The miniJPAS survey quasar selection II: Machine learning classification with photometric measurements and uncertainties