Recursive Bayesian estimation of regularized and irregular quasar light curves

Hanif, Ayub; Protopapas, Pavlos

doi:10.1093/mnras/stv004

Cited by 4 publications

(2 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…C-ARMA models may be useful in comparing the synthetic spectra and light curves produced by numerical models of the accretion disk and winds to observational results. Moving forward, more sophisticated techniques such as the particle filter may be able to directly infer the underlying non-linear stochastic differential equations governing variability without the need for linearization (Hanif & Protopapas 2015). Tools for performing C-ARMA analysis are provided in the c++ and Python package kālī and can be obtained from https://github.com/AstroVPK/libcarma.…”

Section: Discussionmentioning

confidence: 99%

Extracting information from AGN variability

Kasliwal

Vogeley

Richards

2017

Monthly Notices of the Royal Astronomical Society

View full text Add to dashboard Cite

AGN exhibit rapid, high amplitude stochastic flux variations across the entire electromagnetic spectrum on timescales ranging from hours to years. The cause of this variability is poorly understood. We present a Green's Function-based method for using variability to (1) measure the time-scales on which flux perturbations evolve and (2) characterize the driving flux perturbations. We model the observed light curve of an AGN as a linear differential equation driven by stochastic impulses. We analyze the light curve of the Kepler AGN Zw 229-15 and find that the observed variability behavior can be modeled as a damped harmonic oscillator perturbed by a colored noise process. The model powerspectrum turns over on time-scale 385 d. On shorter timescales, the log-powerspectrum slope varies between 2 and 4, explaining the behavior noted by previous studies. We recover and identify both the 5.6 d and 67 d timescales reported by previous work using the Green's Function of the C-ARMA equation rather than by directly fitting the powerspectrum of the light curve. These are the timescales on which flux perturbations grow, and on which flux perturbations decay back to the steady-state flux level respectively. We make the software package kālī used to study light curves using our method available to the community.

show abstract

Section: Discussionmentioning

confidence: 99%

Extracting information from AGN variability

Kasliwal

Vogeley

Richards

2017

Monthly Notices of the Royal Astronomical Society

View full text Add to dashboard Cite

show abstract

“…Automatic classification of variable stars has received substantial attention in the research community in the last years (Debosscher et al 2007;Wachman et al 2009;Kim et al 2009;Wang et al 2010;Richards et al 2011;Kim et al 2011;Pichara et al 2012a;Bloom et al 2012;Pichara & Protopapas 2013;Kim et al 2014;Nun et al 2014;Masci et al 2014;Hanif & Protopapas 2015;Neff et al 2015;Babu & Mahabal 2015). Achieving a good performance with these methods depends strongly on the way lightcurves are represented.…”

Section: Introductionmentioning

confidence: 99%

Clustering-Based Feature Learning on Variable Stars

Mackenzie

Pichara

Protopapas

2016

ApJ

View full text Add to dashboard Cite

The success of automatic classification of variable stars strongly depends on the lightcurve representation. Usually, lightcurves are represented as a vector of many statistical descriptors designed by astronomers called features. These descriptors commonly demand significant computational power to calculate, require substantial research effort to develop and do not guarantee good performance on the final classification task. Today, lightcurve representation is not entirely automatic; algorithms that extract lightcurve features are designed by humans and must be manually tuned up for every survey. The vast amounts of data that will be generated in future surveys like LSST mean astronomers must develop analysis pipelines that are both scalable and automated. Recently, substantial efforts have been made in the machine learning community to develop methods that prescind from expert-designed and manually tuned features for features that are automatically learned from data. In this work we present what is, to our knowledge, the first unsupervised feature learning algorithm designed for variable objects. Our method works by extracting a large number of lightcurve subsequences from a given set of photometric data, which are then clustered to find common local patterns in the time series. Representatives of these common patterns, called exemplars, are then used to transform lightcurves of a labeled set into a new representation that can then be used to train an automatic classifier. The proposed algorithm learns the features from both labeled and unlabeled lightcurves, overcoming the bias generated when the learning process is done only with labeled lightcurves. We test our method on MACHO and OGLE datasets; the results show that the classification performance we achieve is as good and in some cases better than the performance achieved using traditional statistical features, while the computational cost is significantly lower. With these promising results, we believe that our method constitutes a significant step towards the automatization of the lightcurve classification pipeline.

show abstract

Data science methodology for time-delay estimation and data preprocessing of the time-delay challenge

Leon-Anaya

Cuevas-Tello

Valenzuela

et al. 2023

Monthly Notices of the Royal Astronomical Society

View full text Add to dashboard Cite

The vast amount of astronomical information available over this decade has far exceeded the last century. The heterogeneity of the data and the overwhelming magnitude surpassed the possibilities and capabilities to perform manual analysis. As a consequence, new techniques have been developed and different strategies have also been amalgamated such as data science and data mining in order to carry out a more in-depth and exhaustive analysis in search of extraction of knowledge in data. This paper introduces a data science methodology that consists of successive stages, being the core of this proposal the step of data preprocessing, with the aim to reduce the complexity of the analysis and being able to uncover hidden knowledge in the data. The proposed methodology was tested on a set of data consisting of artificial light curves that try to mimic the behaviour of the strong gravitational lens phenomenon supplied by the Time Delay Challenge 1 (TDC1). Under the data science methodology, diverse statistical methods were implemented for data analysis, and cross-correlation and dispersion methods were applied for time delay estimation of strong lensing systems. With this methodology, we obtained time delay estimations from TDC1 dataset and we compared it with previous results reported by the COSmological MOnitoring of GRAvItational Lenses project (COSMOGRAIL). The empirical evidence leads us to conclude that with the proposed methodology, we achieve greater accuracy in estimating time delays in contrast with estimations made with raw data.

show abstract

Recursive Bayesian estimation of regularized and irregular quasar light curves

Cited by 4 publications

References 36 publications

Extracting information from AGN variability

Extracting information from AGN variability

Clustering-Based Feature Learning on Variable Stars

Data science methodology for time-delay estimation and data preprocessing of the time-delay challenge

Contact Info

Product

Resources

About