-
A quantitative Mc Diarmid's inequality for geometrically ergodic Markov chains
Authors:
Antoine Havet,
Matthieu Lerasle,
Eric Moulines,
Elodie Vernet
Abstract:
We state and prove a quantitative version of the bounded difference inequality for geometrically ergodic Markov chains.
Our proof uses the same martingale decomposition as \cite{MR3407208} but, compared to this paper, the exact coupling argument is modified to fill a gap between the strongly aperiodic case and the general aperiodic case.
We state and prove a quantitative version of the bounded difference inequality for geometrically ergodic Markov chains.
Our proof uses the same martingale decomposition as \cite{MR3407208} but, compared to this paper, the exact coupling argument is modified to fill a gap between the strongly aperiodic case and the general aperiodic case.
△ Less
Submitted 5 July, 2019;
originally announced July 2019.
-
A Bayesian nonparametric approach for generalized Bradley-Terry models in random environment
Authors:
Sylvain Le Corff,
Matthieu Lerasle,
Elodie Vernet
Abstract:
This paper deals with the estimation of the unknown distribution of hidden random variables from the observation of pairwise comparisons between these variables. This problem is inspired by recent developments on Bradley-Terry models in random environment since this framework happens to be relevant to predict for instance the issue of a championship from the observation of a few contests per team.…
▽ More
This paper deals with the estimation of the unknown distribution of hidden random variables from the observation of pairwise comparisons between these variables. This problem is inspired by recent developments on Bradley-Terry models in random environment since this framework happens to be relevant to predict for instance the issue of a championship from the observation of a few contests per team. This paper provides three contributions on a Bayesian nonparametric approach to solve this problem. First, we establish contraction rates of the posterior distribution. We also propose a Markov Chain Monte Carlo algorithm to approximately sample from this posterior distribution inspired from a recent Bayesian nonparametric method for hidden Markov models. Finally, the performance of this algorithm are appreciated by comparing predictions on the issue of a championship based on the actual values of the teams and those obtained by sampling from the estimated posterior distribution.
△ Less
Submitted 24 August, 2018;
originally announced August 2018.
-
Efficient semiparametric estimation and model selection for multidimensional mixtures
Authors:
Elisabeth Gassiat,
Judith Rousseau,
Elodie Vernet
Abstract:
In this paper, we consider nonparametric multidimensional finite mixture models and we are interested in the semiparametric estimation of the population weights. Here, the i.i.d. observations are assumed to have at least three components which are independent given the population. We approximate the semiparametric model by projecting the conditional distributions on step functions associated to so…
▽ More
In this paper, we consider nonparametric multidimensional finite mixture models and we are interested in the semiparametric estimation of the population weights. Here, the i.i.d. observations are assumed to have at least three components which are independent given the population. We approximate the semiparametric model by projecting the conditional distributions on step functions associated to some partition. Our first main result is that if we refine the partition slowly enough, the associated sequence of maximum likelihood estimators of the weights is asymptotically efficient, and the posterior distribution of the weights, when using a Bayesian procedure, satisfies a semiparametric Bernstein von Mises theorem. We then propose a cross-validation like procedure to select the partition in a finite horizon. Our second main result is that the proposed procedure satisfies an oracle inequality. Numerical experiments on simulated data illustrate our theoretical results.
△ Less
Submitted 13 December, 2017; v1 submitted 19 July, 2016;
originally announced July 2016.
-
Non Parametric Hidden Markov Models with Finite State Space: Posterior Concentration Rates
Authors:
Elodie Vernet
Abstract:
The use of non parametric hidden Markov models with finite state space is flourishing in practice while few theoretical guarantees are known in this framework. Here, we study asymptotic guarantees for these models in the Bayesian framework. We obtain posterior concentration rates with respect to the $L_1$-norm on joint marginal densities of consecutive observations in a general theorem. We apply t…
▽ More
The use of non parametric hidden Markov models with finite state space is flourishing in practice while few theoretical guarantees are known in this framework. Here, we study asymptotic guarantees for these models in the Bayesian framework. We obtain posterior concentration rates with respect to the $L_1$-norm on joint marginal densities of consecutive observations in a general theorem. We apply this theorem to two cases and obtain minimax concentration rates. We consider discrete observations with emission distributions distributed from a Dirichlet process and continuous observations with emission distributions distributed from Dirichlet process mixtures of Gaussian distributions.
△ Less
Submitted 27 November, 2015;
originally announced November 2015.
-
Posterior consistency for nonparametric hidden Markov models with finite state space
Authors:
Elodie Vernet
Abstract:
In this paper we study posterior consistency for different topologies on the parameters for hidden Markov models with finite state space. We first obtain weak and strong posterior consistency for the marginal density function of finitely many consecutive observations. We deduce posterior consistency for the different components of the parameter. We also obtain posterior consistency for marginal sm…
▽ More
In this paper we study posterior consistency for different topologies on the parameters for hidden Markov models with finite state space. We first obtain weak and strong posterior consistency for the marginal density function of finitely many consecutive observations. We deduce posterior consistency for the different components of the parameter. We also obtain posterior consistency for marginal smoothing distributions in the discrete case. We finally apply our results to independent emission probabilities, translated emission probabilities and discrete HMMs, under various types of priors.
△ Less
Submitted 8 July, 2014; v1 submitted 13 November, 2013;
originally announced November 2013.
-
Sky coverage for Layer Oriented MCAO: a detailed analytical and numerical study
Authors:
Carmelo Arcidiacono,
Emiliano Diolaiti,
Roberto Ragazzoni,
Jacopo Farinato,
Elise Vernet
Abstract:
One of the key-points for the future developments of the multiconjugate adaptive optics for the astronomy is the availability of the correction for a large fraction of the sky. The sky coverage represents one of the limits of the existing single reference adaptive optics system. Multiconjugate adaptive optics allows to overcome the limitations due to the small corrected field of view and the Lay…
▽ More
One of the key-points for the future developments of the multiconjugate adaptive optics for the astronomy is the availability of the correction for a large fraction of the sky. The sky coverage represents one of the limits of the existing single reference adaptive optics system. Multiconjugate adaptive optics allows to overcome the limitations due to the small corrected field of view and the Layer Oriented approach, in particular by its Multiple Field of View version, increases the number of possible references using also very faint stars to guide the adaptive systems. In this paper we study the sky coverage problem in the Layer Oriented case, using both numerical and analytical approaches. Taking into account a star catalogue and a star luminosity distribution function we run a lot of numerical simulation sequences using the Layer Oriented Simulation Tool (LOST). Moreover we perform for several cases a detailed optimization procedure and a relative full simulation in order to achieve better performance for the considered system in those particular conditions. In this way we can retrieve a distribution of numerically simulated cases that allows computing the sky coverage with respect to a performance parameter as the Strehl Ratio and to the scientific field size.
△ Less
Submitted 1 December, 2009; v1 submitted 25 November, 2009;
originally announced November 2009.
-
Layer Oriented Wavefront sensor for MAD on Sky operations
Authors:
Carmelo Arcidiacono,
Matteo Lombini,
Roberto Ragazzoni,
Jacopo Farinato,
Emiliano Diolaiti,
Andrea Baruffolo,
Paolo Bagnara,
Giorgia Gentile,
Laura Schreiber,
Enrico Marchetti,
Johann Kolb,
Sebastien Tordo,
Robert Donaldson,
Christian Soenke,
Sylvain Oberti,
Enrico Fedrigo,
Elise Vernet,
Norbert Hubin
Abstract:
The Multiconjugate Adaptive optics Demonstrator (MAD) has successfully demonstrated on sky both Star Oriented (SO) and Layer Oriented (LO) multiconjugate adaptive optics techniques. While SO has been realized using 3 Shack-Hartmann wavefront sensors (WFS), we designed a multi-pyramid WFS for the LO. The MAD bench accommodates both WFSs and a selecting mirror allows choosing which sensor to use.…
▽ More
The Multiconjugate Adaptive optics Demonstrator (MAD) has successfully demonstrated on sky both Star Oriented (SO) and Layer Oriented (LO) multiconjugate adaptive optics techniques. While SO has been realized using 3 Shack-Hartmann wavefront sensors (WFS), we designed a multi-pyramid WFS for the LO. The MAD bench accommodates both WFSs and a selecting mirror allows choosing which sensor to use. In the LO approach up to 8 pyramids can be placed on as many reference stars and their light is co-added optically on two different CCDs conjugated at ground and to an high layer. In this paper we discuss LO commissioning phase and on sky operations.
△ Less
Submitted 25 November, 2009;
originally announced November 2009.
-
The Blue Straggler population in the globular cluster M53 (NGC5024): a combined HST, LBT, CFHT study
Authors:
G. Beccari,
B. Lanzoni,
F. R. Ferraro,
L. Pulone,
M. Bellazzini,
F. Fusi Pecci,
R. T. Rood,
E. Giallongo,
R. Ragazzoni,
A. Grazian,
A. Baruffolo,
N. Bouche,
P. Buschkamp,
C. De Santis,
E. Diolaiti,
A. Di Paola,
J. Farinato,
A. Fontana,
S. Gallozzi,
F. Gasparo,
G. Gentile,
F. Pasian,
F. Pedichini,
R. Smareglia,
R. Speziali
, et al. (2 additional authors not shown)
Abstract:
We used a proper combination of multiband high-resolution and wide field multi-wavelength observations collected at three different telescopes (HST, LBT and CFHT) to probe Blue Straggler Star (BSS) populations in the globular cluster M53. Almost 200 BSS have been identified over the entire cluster extension. The radial distribution of these stars has been found to be bimodal (similarly to that o…
▽ More
We used a proper combination of multiband high-resolution and wide field multi-wavelength observations collected at three different telescopes (HST, LBT and CFHT) to probe Blue Straggler Star (BSS) populations in the globular cluster M53. Almost 200 BSS have been identified over the entire cluster extension. The radial distribution of these stars has been found to be bimodal (similarly to that of several other clusters) with a prominent dip at ~60'' (~2 r_c) from the cluster center. This value turns out to be a factor of two smaller than the radius of avoidance (r_avoid, the radius within which all the stars of ~1.2 M_sun have sunk to the core because of dynamical friction effects in an Hubble time). While in most of the clusters with a bimodal BSS radial distribution, r_avoid has been found to be located in the region of the observed minimum, this is the second case (after NGC6388) where this discrepancy is noted. This evidence suggests that in a few clusters the dynamical friction seems to be somehow less efficient than expected.
We have also used this data base to construct the radial star density profile of the cluster: this is the most extended and accurate radial profile ever published for this cluster, including detailed star counts in the very inner region. The star density profile is reproduced by a standard King Model with an extended core (~25'') and a modest value of the concentration parameter (c=1.58). A deviation from the model is noted in the most external region of the cluster (at r>6.5' from the center). This feature needs to be further investigated in order to address the possible presence of a tidal tail in this cluster.
△ Less
Submitted 19 February, 2008;
originally announced February 2008.
-
The performance of the blue prime focus Large Binocular Camera at the Large Binocular Telescope
Authors:
E. Giallongo,
R. Ragazzoni,
A. Grazian,
A. Baruffolo,
G. Beccari,
C. De Santis,
E. Diolaiti,
A. Di Paola,
J. Farinato,
A. Fontana,
S. Gallozzi,
F. Gasparo,
G. Gentile,
R. Green,
J. Hill,
O. Kuhn,
F. Pasian,
F. Pedichini,
M. Radovich,
P. Salinari,
R. Smareglia,
R. Speziali,
V. Testa,
D. Thompson,
E. Vernet
, et al. (1 additional authors not shown)
Abstract:
We present the characteristics and some early scientific results of the first instrument at the Large Binocular Telescope (LBT), the Large Binocular Camera (LBC). Each LBT telescope unit will be equipped with similar prime focus cameras. The blue channel is optimized for imaging in the UV-B bands and the red channel for imaging in the VRIz bands. The corrected field-of-view of each camera is app…
▽ More
We present the characteristics and some early scientific results of the first instrument at the Large Binocular Telescope (LBT), the Large Binocular Camera (LBC). Each LBT telescope unit will be equipped with similar prime focus cameras. The blue channel is optimized for imaging in the UV-B bands and the red channel for imaging in the VRIz bands. The corrected field-of-view of each camera is approximately 30 arcminutes in diameter, and the chip area is equivalent to a 23x23 arcmin2 field. In this paper we also present the commissioning results of the blue channel. The scientific and technical performance of the blue channel was assessed by measurement of the astrometric distortion, flat fielding, ghosts, and photometric calibrations. These measurements were then used as input to a data reduction pipeline applied to science commissioning data. The measurements completed during commissioning show that the technical performance of the blue channel is in agreement with original expectations. Since the red camera is very similar to the blue one we expect similar performance from the commissioning that will be performed in the following months in binocular configuration. Using deep UV image, acquired during the commissioning of the blue camera, we derived faint UV galaxy-counts in a ~500 sq. arcmin. sky area to U(Vega)=26.5. These galaxy counts imply that the blue camera is the most powerful UV imager presently available and in the near future in terms of depth and extent of the field-of-view. We emphasize the potential of the blue camera to increase the robustness of the UGR multicolour selection of Lyman break galaxies at redshift z~3.
△ Less
Submitted 13 March, 2008; v1 submitted 9 January, 2008;
originally announced January 2008.
-
A near-ultraviolet view of the Inner Region of M31 with the Large Binocular Telescope
Authors:
G. Beccari,
M. Bellazzini,
G. Clementini,
L. Federici,
F. Fusi Pecci,
S. Galleti,
P. Montegriffo,
E. Giallongo,
R. Ragazzoni,
A. Grazian,
A. Baruffolo,
C. De Santis,
E. Diolaiti,
A. Di Paola,
J. Farinato,
A. Fontana,
S. Gallozzi,
F. Gasparo,
G. Gentile,
R. Green,
J. Hill,
O. Kuhn,
N. Menci,
F. Pasian F. Pedichini,
R. Smareglia
, et al. (5 additional authors not shown)
Abstract:
We present a 900 sec, wide-field U image of the inner region of the Andromeda galaxy obtained during the commissioning of the blue channel of the Large Binocular Camera mounted on the prime focus of the Large Binocular Telescope. Relative photometry and absolute astrometry of individual sources in the image was obtained along with morphological parameters aimed at discriminating between stars an…
▽ More
We present a 900 sec, wide-field U image of the inner region of the Andromeda galaxy obtained during the commissioning of the blue channel of the Large Binocular Camera mounted on the prime focus of the Large Binocular Telescope. Relative photometry and absolute astrometry of individual sources in the image was obtained along with morphological parameters aimed at discriminating between stars and extended sources, e.g. globular clusters. The image unveils the near-ultraviolet view of the inner ring of star formation recently discovered in the infrared by the Spitzer Space Telescope and shows in great detail the fine structure of the dust lanes associated with the galaxy inner spiral arms. The capabilities of the blue channel of the Large Binocular Camera at the Large Binocular Telescope (LBC-Blue) are probed by direct comparison with ultraviolet GALEX observations of the same region in M31. We discovered 6 new candidate stellar clusters in this high-background region of M31. We also recovered 62 bona-fide globulars and 62 previously known candidates from the Revised Bologna Catalogue of the M31 globular clusters, and firmly established the extended nature of 19 of them.
△ Less
Submitted 9 October, 2007;
originally announced October 2007.
