| 1. |
- Dainotti, Maria Giovanna, et al.
(författare)
-
Predicting the Redshift of γ-Ray-loud AGNs Using Supervised Machine Learning
- 2021
-
Ingår i: Astrophysical Journal. - : American Astronomical Society. - 0004-637X .- 1538-4357. ; 920:2
-
Tidskriftsartikel (refereegranskat)abstract
- Active galactic nuclei (AGNs) are very powerful galaxies characterized by extremely bright emissions coming from their central massive black holes. Knowing the redshifts of AGNs provides us with an opportunity to determine their distance to investigate important astrophysical problems, such as the evolution of the early stars and their formation, along with the structure of early galaxies. The redshift determination is challenging because it requires detailed follow-up of multiwavelength observations, often involving various astronomical facilities. Here we employ machine-learning algorithms to estimate redshifts from the observed γ-ray properties and photometric data of γ-ray-loud AGNs from the Fourth Fermi-LAT Catalog. The prediction is obtained with the Superlearner algorithm using a LASSO-selected set of predictors. We obtain a tight correlation, with a Pearson correlation coefficient of 71.3% between the inferred and observed redshifts and an average Δz norm = 11.6 10-4. We stress that, notwithstanding the small sample of γ-ray-loud AGNs, we obtain a reliable predictive model using Superlearner, which is an ensemble of several machine-learning models.
|
|
| 2. |
- Hallgren, Jonas
(författare)
-
Continuous time Graphical Models and Decomposition Sampling
- 2015
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Two topics in temporal graphical probabilistic models are studied. The topics are treated in separate papers, both with applications in finance. The first paper study inference in dynamic Bayesian networks using Monte Carlo methods. A new method for sampling random variables is proposed. The method divides the sample space into subspaces. This allows the sampling to be done in parallel with independent and distinct sampling methods on the subspaces. The methodology is demonstrated on a volatility model and some toy examples with promising results. The second paper treats probabilistic graphical models in continuous time —a class of models with the ability to express causality. Tools for inference in these models are developed and employed in the design of a causality measure. The framework is used to analyze tick-by-tick data from the foreign exchange market.
|
|
| 3. |
- Jiang, Wei, et al.
(författare)
-
Adaptive Bayesian SLOPE : Model Selection With Incomplete Data
- 2022
-
Ingår i: Journal of Computational and Graphical Statistics. - : Informa UK Limited. - 1061-8600 .- 1537-2715. ; 31:1, s. 113-137
-
Tidskriftsartikel (refereegranskat)abstract
- We consider the problem of variable selection in high-dimensional settings with missing observations among the covariates. To address this relatively understudied problem, we propose a new synergistic procedure—adaptive Bayesian SLOPE with missing values—which effectively combines SLOPE (sorted l 1 regularization) with the spike-and-slab LASSO (SSL) and is accompanied by an efficient stochastic approximation of expected maximization (SAEM) algorithm to handle missing data. Similarly as in SSL, the regression coefficients are regarded as arising from a hierarchical model consisting of two groups: the spike for the inactive and the slab for the active. However, instead of assigning independent spike and slab Laplace priors for each covariate, here we deploy a joint SLOPE “spike-and-slab” prior which takes into account the ordering of coefficient magnitudes in order to control for false discoveries. We position our approach within a Bayesian framework which allows for simultaneous variable selection and parameter estimation while handling missing data. Through extensive simulations, we demonstrate satisfactory performance in terms of power, false discovery rate (FDR) and estimation bias under a wide range of scenarios including complete data and existence of missingness. Finally, we analyze a real dataset consisting of patients from Paris hospitals who underwent severe trauma, where we show competitive performance in predicting platelet levels. Our methodology has been implemented in C++ and wrapped into open source R programs for public use. Supplemental files for this article are available online.
|
|
