PERIODICITY DETECTION IN AGN WITH THE BOOSTED TREE METHOD

We apply a machine learning algorithm called XGBoost to explore the periodicity of two radio sources: PKS 1921-293 (OV 236) and PKS 2200+420 (BL Lac), both radio frequency datasets obtained from University of Michigan Radio Astronomy Observatory (UMRAO), at 4.8 GHz, 8.0 GHz, and 14.5 GHz, between 1969 to 2012. From this methods, we find that the XGBoost provides the opportunity to use a machine learning based methodology on radio datasets and to extract information with strategies quite different from those traditionally used to treat time series, as well as to obtain periodicity through the classification of recurrent events. The results were compared with other methods that examined the same datasets and exhibit a good agreement with them.

Periodicity of quasar and galaxy redshift

Context. An approach for studying the large-scale structure of the Universe lies in the detection and analysis of periodicity of redshift values of extragalactic objects, galaxies, and quasi stellar objects (QSO), in particular. Moreover, the hypothesis of the existence of multiple periodicities in the redshift distributions deserves exploration. The task is compounded by the presence of confounding effects and measurement noise. Aims. Studies of periodicity detection in redshift values of extragalactic objects obtained from the Sloan Digital Sky Survey (SDSS) have been conducted in the past, largely based on the Fourier transform. The present study aims to revisit the same thing using the singular value decomposition (SVD) as the primary tool. Methods. Periodicity detection and the determination of the fundamental period have been performed using a standard spectral approach as well as a SVD-based approach for a variety of simulated datasets. The analysis of the quasar redshift data from DR12 and galaxy redshift dataset of DR10 from SDSS data has been carried out. Results. A wide range of periodicities are observed in different redshift ranges of the quasar datasets. For redshifts greater than 0.03, a period length of 0.2094 was determined while periodicities of 0.1210 and 0.0654 were obtained for redshift ranges (0.03, 1) and (3, 4), respectively. Twin periodicities of 0.1153 and 0.0807 were obtained for the redshift range (1, 3). Determining the ranges to be examined has been done based on histogram computation; the binwidths of which have been obtained by employing a kernel density estimation. The redshift sequence for the galaxy samples exhibits a somewhat different nature, but still contains periodic components. Twin periodicities of 0.0056 and 0.0079 were observed for a redshift range greater than 0.03. Conclusions. Galaxy and quasar redshift values form sequences, which are not only discrete in amplitude but also contain periodic components. The superiority of the singular value decomposition method over the spectral estimation approach, in redshift periodicity analysis especially from the point of view of robustness, is demonstrated through simulations. The existence of periodicity for quasar and galaxy families is thus firmly established, lending further support to the Hoyle-Narlikar variable mass theory.

Signal periodicity detection using Ramanujan subspace projection

Signal periodic decomposition and periodic estimation are two crucial problems in the signal processing domain. Due to its significance, the applications have been extended to fields like periodic sequence analysis of biomolecules, stock market predictions, speech signal processing, and musical pitch analysis. The recently proposed Ramanujan sums (RS) based transforms are very useful in analysing the periodicity of signals. This paper proposes a method for periodicity detection of signals with multiple periods based on autocorrelation and Ramanujan subspace projection with low computational complexity. The proposed method is compared with other signal periodicity detection methods and the results show that the proposed method detects the signal period correctly in less time.

Two-dimensional correlation analysis of periodicity in active galactic nuclei time series

The active galactic nuclei (AGN) are among the most powerful sources with an inherent, pronounced and random variation of brightness. The randomness of their time series is so subtle as to blur the border between aperiodic fluctuations and noisy oscillations. This poses challenges to analysing of such time series because neither visual inspection nor pre-exisitng methods can identify well oscillatory signals in them. Thus, there is a need for an objective method for periodicity detection. Here we review our a new data analysis method that combines a two-dimensional correlation (2D) of time series with the powerful methods of Gaussian processes. To demonstrate the utility of this technique, we apply it to two example problems which were not exploited enough: damped rednoised artificial time series mimicking AGN time series and newly published observed time series of changing look AGN (CL AGN) NGC 3516. The method successfully detected periodicities in both types of time series. Identified periodicity of ~4 yr in NGC 3516 allows us to speculate that if the thermal instability formed in its accretion disc (AD) on a time scale resembling detected periodicity then AD radius could be ~0.0024 pc.