Abstract:We present the construction of a novel time-domain signature extraction methodology and the development of a supporting supervised pattern detection algorithm. We focus on the targeted identification of eclipsing binaries that demonstrate a feature known as the O'Connell effect. Our proposed methodology maps stellar variable observations to a new representation known as distribution fields (DFs). Given this novel representation, we develop a metric learning technique directly on the DF space that is capable of… Show more
“…SSMM itself is an effective feature for discriminating variable star types as shown by Johnston and Peter (2017). Similarly, DF has been shown to be a valuable feature for discriminating time domain signatures, see Helfer et al (2015) and Johnston et al (2019).…”
Section: Feature Extractionmentioning
confidence: 95%
“…Although our initial goal is variable star identification, given a separate set of features this method could be applied to other astroinformatics problems (i.e., image classification for galaxies, spectral identification for stars or comets, etc.). While we demonstrate the classifier has a multi-class classification design, which is common in the astroinformatics references we have provided, the design here can easily be transformed into a one-vs-all design (Johnston and Oluseyi 2017) for the purposes of generating a detector or classifier designed specifically to a user's needs (Johnston et al 2019).…”
Section: Theory and Designmentioning
confidence: 99%
“…Prior studies have initially addressed the potential of using metric learning as a means for classification of variable stars (Johnston et al 2019). Metric learning has a number of benefits that are advantageous to the astronomer:…”
Section: Classification and Metric Learningmentioning
confidence: 99%
“…Presented here is a methodology that addresses these issues using a combination of new features and advanced classifiers designs. Two novel transforms, Slotted Symbolic Markov Model (SSMM, Johnston and Peter 2017) and Distribution Fields (DF, Johnston et al 2019), are used to generate viable feature spaces for the classification of variable stars. SSMM requires no phasing of the time domain data but still provides a feature that is shape based, DF allows for the consideration of the whole phased waveform without additional picking and choosing of metrics from the waveform (i.e., see Helfer et al 2015).…”
Context. Comprehensive observations of variable stars can include time domain photometry in a multitude of filters, spectroscopy, estimates of color (e.g. U-B), etc. When the objective is to classify variable stars, traditional machine learning techniques distill these various representations (or views) into a single feature vector and attempt to discriminate among desired categories. Aims. In this work, we propose an alternative approach that inherently leverages multiple views of the same variable star. Methods. Our multi-view metric learning framework enables robust characterization of star categories by directly learning to discriminate in a multi-faceted feature space, thus, eliminating the need to combine feature representations prior to fitting the machine learning model. We also demonstrate how to extend standard multi-view learning, which employs multiple vectorized views, to the matrix-variate case which allows very novel variable star signature representations.Results. The performance of our proposed methods is evaluated on the UCR Starlight and LINEAR datasets. Both the vector and matrix-variate versions of our multi-view learning framework perform favorably -demonstrating the ability to discriminate variable star categories..
“…SSMM itself is an effective feature for discriminating variable star types as shown by Johnston and Peter (2017). Similarly, DF has been shown to be a valuable feature for discriminating time domain signatures, see Helfer et al (2015) and Johnston et al (2019).…”
Section: Feature Extractionmentioning
confidence: 95%
“…Although our initial goal is variable star identification, given a separate set of features this method could be applied to other astroinformatics problems (i.e., image classification for galaxies, spectral identification for stars or comets, etc.). While we demonstrate the classifier has a multi-class classification design, which is common in the astroinformatics references we have provided, the design here can easily be transformed into a one-vs-all design (Johnston and Oluseyi 2017) for the purposes of generating a detector or classifier designed specifically to a user's needs (Johnston et al 2019).…”
Section: Theory and Designmentioning
confidence: 99%
“…Prior studies have initially addressed the potential of using metric learning as a means for classification of variable stars (Johnston et al 2019). Metric learning has a number of benefits that are advantageous to the astronomer:…”
Section: Classification and Metric Learningmentioning
confidence: 99%
“…Presented here is a methodology that addresses these issues using a combination of new features and advanced classifiers designs. Two novel transforms, Slotted Symbolic Markov Model (SSMM, Johnston and Peter 2017) and Distribution Fields (DF, Johnston et al 2019), are used to generate viable feature spaces for the classification of variable stars. SSMM requires no phasing of the time domain data but still provides a feature that is shape based, DF allows for the consideration of the whole phased waveform without additional picking and choosing of metrics from the waveform (i.e., see Helfer et al 2015).…”
Context. Comprehensive observations of variable stars can include time domain photometry in a multitude of filters, spectroscopy, estimates of color (e.g. U-B), etc. When the objective is to classify variable stars, traditional machine learning techniques distill these various representations (or views) into a single feature vector and attempt to discriminate among desired categories. Aims. In this work, we propose an alternative approach that inherently leverages multiple views of the same variable star. Methods. Our multi-view metric learning framework enables robust characterization of star categories by directly learning to discriminate in a multi-faceted feature space, thus, eliminating the need to combine feature representations prior to fitting the machine learning model. We also demonstrate how to extend standard multi-view learning, which employs multiple vectorized views, to the matrix-variate case which allows very novel variable star signature representations.Results. The performance of our proposed methods is evaluated on the UCR Starlight and LINEAR datasets. Both the vector and matrix-variate versions of our multi-view learning framework perform favorably -demonstrating the ability to discriminate variable star categories..
“…Considerable efforts have gone into using machine learning to classify light curves from large ground-based surveys (e.g., Carrasco-Davis et al 2019;Tsang & Schultz 2019;Johnston et al 2019a;Cabral et al 2020;Hosenie et al 2020;Jamal & Bloom 2020;Szklenár et al 2020;Bassi et al 2021;Zhang & Bloom 2021). Such techniques have also been applied to light curves from NASA's Kepler and K2 missions (e.g., Blomme et al 2010Blomme et al , 2011Debosscher et al 2011;Bass & Borne 2016;Armstrong et al 2016;Hon et al 2017Hon et al , 2018bJohnston et al 2019b;Kgoadi et al 2019;Le Saux et al 2019;Giles & Walkowicz 2020;Kuszlewicz et al 2020;Audenaert et al 2021;Paul & Chattopadhyay 2022).…”
With the availability of large-scale surveys like Kepler and TESS, there is a pressing need for automated methods to classify light curves according to known classes of variable stars. We introduce a new algorithm for classifying light curves that compares 7000 time-series features to find those which most effectively classify a given set of light curves. We apply our method to Kepler light curves for stars with effective temperatures in the range 6500-10,000 K. We show that the sample can be meaningfully represented in an interpretable five-dimensional feature space that separates seven major classes of light curves (𝛿 Scuti stars, 𝛾 Doradus stars, RR Lyrae stars, rotational variables, contact eclipsing binaries, detached eclipsing binaries, and non-variables). We achieve a balanced classification accuracy of 82% on an independent test set of Kepler stars using a Gaussian mixture model classifier. We use our method to classify 12,000 Kepler light curves from Quarter 9 and provide a catalogue of the results. We further outline a confidence heuristic based on probability density with which to search our catalogue, and extract candidate lists of correctly-classified variable stars.
Context. It is now clear that binarity plays a crucial role in many aspects of planetary nebulae (PNe), particularly the striking morphologies that they show. To date, there are ∼ 60 bCSPNe known. However, both theory and observation indicates that this represents only the tip of the iceberg, with the Galactic PN population hosting orders of magnitude more. Aims. We are involved in a search for new bCSPNe to enhance the statistical validation of the key role of binarity in the formation and shaping of PNe. New discoveries of bCSPNe and their characterization have important implications not only in understanding PN evolution but also in understanding binary evolution and the poorly-understood common-envelope phase. Methods. We used data from the TESS satellite to search for variability in the eight CSPNe that belong to the two-minute cadence preselected targets in Cycle 1, which have available pipeline-extracted light curves. We identified strong periodicities and analysed them in the context of the binary scenario. Results. All the CSPNe but one (Abell 15) show clear signs of periodic variability in TESS. The cause of this variability can be attributed to different effects, some of them requiring the presence of a companion star. We find simple sinusoidal modulations in several of the systems, compatible to irradiation effects. In addition, two of the central stars (PG 1034+001 and NGC 5189) also show photometric variations due to ellipsoidal variations and other signs of variability probably caused by star spots and/or relativistic Doppler-beaming. Especially interesting is the case of the well-studied Helix Nebula, in which we constructed a series of binary models to explain the modulations we see in the light curve. We find that the variability constrains the possible companion to be very low-mass main-sequence star or sub-stellar object. We also identify with a great detail the individual pulsation frequencies of NGC 246.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.