Search for: All records

We reanalysed the c-type RR Lyrae star BE Dor (MACHO 5.4644.8, OGLE-LMC-RRLYR-06002) that had been discovered to show cyclic period changes. The photometric data of several sky surveys (DASCH, MACHO, OGLE, ASAS-SN, and TESS) were used for analyses. The O − C diagram and pulsation period obtained from Fourier analysis show significant period modulations in BE Dor. However, different from the previous viewpoint, the changes are quasi-periodic and abrupt. Therefore, the light-traveltime effect caused by the companion motion cannot explain the changes. Noting a same subtype star KIC 9453114 with similar phenomena has a high macroturbulent velocity, and the degree of O − C changes seem to be positively correlated with these velocities, we consider that the mechanism leading to period modulation should be caused by the interaction between turbulent convection and magnetic field activity in the ionization zone, i.e. the viewpoint of Stothers. It may not explain the general Blazhko effect but should explain such period modulations in BE Dor and those other c-type RR Lyrae stars. We hope our discoveries and viewpoints can provide some information and inspiration for relevant research.

 

Stochastic gradient Langevin dynamics (SGLD) and stochastic gradient Hamiltonian Monte Carlo (SGHMC) are two popular Markov Chain Monte Carlo (MCMC) algorithms for Bayesian inference that can scale to large datasets, allowing to sample from the posterior distribution of the parameters of a statistical model given the input data and the prior distribution over the model parameters. However, these algorithms do not apply to the decentralized learning setting, when a network of agents are working collaboratively to learn the parameters of a statistical model without sharing their individual data due to privacy reasons or communication constraints. We study two algorithms: Decentralized SGLD (DE-SGLD) and Decentralized SGHMC (DE-SGHMC) which are adaptations of SGLD and SGHMC methods that allow scaleable Bayesian inference in the decentralized setting for large datasets. We show that when the posterior distribution is strongly log-concave and smooth, the iterates of these algorithms converge linearly to a neighborhood of the target distribution in the 2-Wasserstein distance if their parameters are selected appropriately. We illustrate the efficiency of our algorithms on decentralized Bayesian linear regression and Bayesian logistic regression problems