Arnak Dalalyan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Arnak Dalalyan has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Renaud Keriven
Université Paris-Est
2 shared publications - Alexandre B. Tsybakov
CREST-ENSAE
1 shared publication B. S. Nahapetian
National Academy of Sciences of Armenia
1 shared publication- Jérôme Courchay
Université Paris-Est
1 shared publication - Laëtitia Comminges
LIGM/IMAGINE CREST/GENES ENSAE
1 shared publication - Laëtitia Comminges
Université Paris Est/ ENPC LIGM IMAGINE
1 shared publication
Latest Publications
- Open Access
Minimax testing of a composite null hypothesis defined via a quadratic functional in the model of regression
We consider the problem of testing a particular type of composite null hypothesis under a nonparametric multivariate regression model. For a given quadratic functional $Q$, the null hypothesis states that the regression function $f$ satisfies the constraint $Q[f]=0$, while the alternative corresponds to the functions for which $Q[f]$ is bounded away from zero. On the one hand, we provide minimax rates of testing and the exact separation constants, along with a sharp-optimal testing procedure, for diagonal and nonnegative quadratic functionals. We consider smoothness classes of ellipsoidal form and check that our conditions are fulfilled in the particular case of ellipsoids corresponding to anisotropic Sobolev classes. In this case, we present a closed form of the minimax rate and the separation constant. On the other hand, minimax rates for quadratic functionals which are neither positive nor negative makes appear two different regimes: "regular" and "irregular". In the "regular" case, the minimax rate is equal to $n^{-1/4}$ while in the "irregular" case, the rate depends on the smoothness class and is slower than in the "regular" case. We apply this to the issue of testing the equality of norms of two functions observed in noisy environments.Open AccessMinimax testing of a composite null hypothesis defined via a quadratic functional in the model of regression
We consider the problem of testing a particular type of composite null hypothesis under a nonparametric multivariate regression model. For a given quadratic functional $Q$, the null hypothesis states that the regression function $f$ satisfies the constraint $Q[f]=0$, while the alternative corresponds to the functions for which $Q[f]$ is bounded away from zero. On the one hand, we provide minimax rates of testing and the exact separation constants, along with a sharp-optimal testing procedure, for diagonal and nonnegative quadratic functionals. We consider smoothness classes of ellipsoidal form and check that our conditions are fulfilled in the particular case of ellipsoids corresponding to anisotropic Sobolev classes. In this case, we present a closed form of the minimax rate and the separation constant. On the other hand, minimax rates for quadratic functionals which are neither positive nor negative makes appear two different regimes: "regular" and "irregular". In the "regular" case, the minimax rate is equal to $n^{-1/4}$ while in the "irregular" case, the rate depends on the smoothness class and is slower than in the "regular" case. We apply this to the issue of testing the equality of norms of two functions observed in noisy environments. - 3 August 2011
- Published by Springer Nature in International Journal of Computer Vision
- Vol. 97 (1) , 71-90
On Camera Calibration with Linear Programming and Loop Constraint Linearization
- 3 August 2011
- Published by Springer Nature in International Journal of Computer Vision
- Vol. 97 (1) , 71-90
On Camera Calibration with Linear Programming and Loop Constraint Linearization
- 9 April 2011
- Published by Springer Nature in Journal of Mathematical Imaging and Vision
- Vol. 43 (1) , 10-23
Robust Estimation for an Inverse Problem Arising in Multiview Geometry
- 9 April 2011
- Published by Springer Nature in Journal of Mathematical Imaging and Vision
- Vol. 43 (1) , 10-23
Robust Estimation for an Inverse Problem Arising in Multiview Geometry
- 1 April 2011
- Published by Allerton Press in Journal of Contemporary Mathematical Analysis (Armenian Academy of Sciences)
- Vol. 46 (2) , 113-119
Description of random fields by means of one-point finite-conditional distributions
The aim of this note is to investigate the relationship between strictly positive random fields on a lattice ℤ ν and the conditional probability measures at one point given the values on a finite subset of the lattice ℤ ν . We exhibit necessary and sufficient conditions for a one-point finite-conditional system to correspond to a unique strictly positive probability measure. It is noteworthy that the construction of the aforementioned probability measure is done explicitly by some simple procedure. Finally, we introduce a condition on the one-point finite conditional system that is sufficient for ensuring the mixing of the underlying random field.
- 1 April 2011
- Published by Allerton Press in Journal of Contemporary Mathematical Analysis (Armenian Academy of Sciences)
- Vol. 46 (2) , 113-119
Description of random fields by means of one-point finite-conditional distributions
The aim of this note is to investigate the relationship between strictly positive random fields on a lattice ℤ ν and the conditional probability measures at one point given the values on a finite subset of the lattice ℤ ν . We exhibit necessary and sufficient conditions for a one-point finite-conditional system to correspond to a unique strictly positive probability measure. It is noteworthy that the construction of the aforementioned probability measure is done explicitly by some simple procedure. Finally, we introduce a condition on the one-point finite conditional system that is sufficient for ensuring the mixing of the underlying random field.
- Open Access
Tight conditions for consistent variable selection in high dimensional nonparametric regression
We address the issue of variable selection in the regression model with very high ambient dimension, i.e., when the number of covariates is very large. The main focus is on the situation where the number of relevant covariates, called intrinsic dimension, is much smaller than the ambient dimension. Without assuming any parametric form of the underlying regression function, we get tight conditions making it possible to consistently estimate the set of relevant variables. These conditions relate the intrinsic dimension to the ambient dimension and to the sample size. The procedure that is provably consistent under these tight conditions is simple and is based on comparing the empirical Fourier coefficients with an appropriately chosen threshold value.
Open AccessTight conditions for consistent variable selection in high dimensional nonparametric regression
We address the issue of variable selection in the regression model with very high ambient dimension, i.e., when the number of covariates is very large. The main focus is on the situation where the number of relevant covariates, called intrinsic dimension, is much smaller than the ambient dimension. Without assuming any parametric form of the underlying regression function, we get tight conditions making it possible to consistently estimate the set of relevant variables. These conditions relate the intrinsic dimension to the ambient dimension and to the sample size. The procedure that is provably consistent under these tight conditions is simple and is based on comparing the empirical Fourier coefficients with an appropriately chosen threshold value.
- Open Access
Second-order asymptotic expansion for a non-synchronous covariation estimator
In this paper, we consider the problem of estimating the covariation of two diffusion processes when observations are subject to non-synchronicity. Building on recent papers \cite{Hay-Yos03, Hay-Yos04}, we derive second-order asymptotic expansions for the distribution of the Hayashi-Yoshida estimator in a fairly general setup including random sampling schemes and non-anticipative random drifts. The key steps leading to our results are a second-order decomposition of the estimator's distribution in the Gaussian set-up, a stochastic decomposition of the estimator itself and an accurate evaluation of the Malliavin covariance. To give a concrete example, we compute the constants involved in the resulting expansions for the particular case of sampling scheme generated by two independent Poisson processes.
Open AccessSecond-order asymptotic expansion for a non-synchronous covariation estimator
In this paper, we consider the problem of estimating the covariation of two diffusion processes when observations are subject to non-synchronicity. Building on recent papers \cite{Hay-Yos03, Hay-Yos04}, we derive second-order asymptotic expansions for the distribution of the Hayashi-Yoshida estimator in a fairly general setup including random sampling schemes and non-anticipative random drifts. The key steps leading to our results are a second-order decomposition of the estimator's distribution in the Gaussian set-up, a stochastic decomposition of the estimator itself and an accurate evaluation of the Malliavin covariance. To give a concrete example, we compute the constants involved in the resulting expansions for the particular case of sampling scheme generated by two independent Poisson processes.
- Open Access
Sparse Regression Learning by Aggregation and Langevin Monte-Carlo
We consider the problem of regression learning for deterministic design and independent random errors. We start by proving a sharp PAC-Bayesian type bound for the exponentially weighted aggregate (EWA) under the expected squared empirical loss. For a broad class of noise distributions the presented bound is valid whenever the temperature parameter $\beta$ of the EWA is larger than or equal to $4\sigma^2$, where $\sigma^2$ is the noise variance. A remarkable feature of this result is that it is valid even for unbounded regression functions and the choice of the temperature parameter depends exclusively on the noise level. Next, we apply this general bound to the problem of aggregating the elements of a finite-dimensional linear space spanned by a dictionary of functions $\phi_1,...,\phi_M$. We allow $M$ to be much larger than the sample size $n$ but we assume that the true regression function can be well approximated by a sparse linear combination of functions $\phi_j$. Under this sparsity scenario, we propose an EWA with a heavy tailed prior and we show that it satisfies a sparsity oracle inequality with leading constant one. Finally, we propose several Langevin Monte-Carlo algorithms to approximately compute such an EWA when the number $M$ of aggregated functions can be large. We discuss in some detail the convergence of these algorithms and present numerical experiments that confirm our theoretical findings.Open AccessSparse Regression Learning by Aggregation and Langevin Monte-Carlo
We consider the problem of regression learning for deterministic design and independent random errors. We start by proving a sharp PAC-Bayesian type bound for the exponentially weighted aggregate (EWA) under the expected squared empirical loss. For a broad class of noise distributions the presented bound is valid whenever the temperature parameter $\beta$ of the EWA is larger than or equal to $4\sigma^2$, where $\sigma^2$ is the noise variance. A remarkable feature of this result is that it is valid even for unbounded regression functions and the choice of the temperature parameter depends exclusively on the noise level. Next, we apply this general bound to the problem of aggregating the elements of a finite-dimensional linear space spanned by a dictionary of functions $\phi_1,...,\phi_M$. We allow $M$ to be much larger than the sample size $n$ but we assume that the true regression function can be well approximated by a sparse linear combination of functions $\phi_j$. Under this sparsity scenario, we propose an EWA with a heavy tailed prior and we show that it satisfies a sparsity oracle inequality with leading constant one. Finally, we propose several Langevin Monte-Carlo algorithms to approximately compute such an EWA when the number $M$ of aggregated functions can be large. We discuss in some detail the convergence of these algorithms and present numerical experiments that confirm our theoretical findings.