Searcharxiv⌕ Search

arXiv subjects

J Dedecker

Publications and source records attributed to J Dedecker.

14 recordsLinked to original sources
arXiv2025

A model-free Screening procedure

In this article, we propose a generic screening method for selecting explanatory variables correlated with the response variable Y . We make no assumptions about the existence of a model that could link Y with a subset of explanatory variables, nor about the distribution of the variables. Our procedure can therefore be described as ''model-free'' and can be applied in a wide range of situations. In order to obtain precise theoretical guarantees (Sure Screening Property and control of the False Positive Rate), we establish a Berry-Esseen type inequality for the studentized statistic of the slope estimator. We illustrate our selection procedure using two simulated examples and a real data set.

math.ST↗
arXiv2025

On Berry Esseen type estimates for randomized Martingales in the non stationary setting

In this paper, we consider partial sums of triangular martingale differences weighted by random variables drawn uniformly on the sphere, and globally independent of the martingale differences. Starting from the so-called principle of conditioning and using some arguments developed by Klartag-Sodin and Bobkov-Chistyakov-G{\"o}tze, we give some upper bounds for the Kolmogorov distance between the distribution of these weighted sums and a Normal distribution. Under some conditions on the conditional variances of the martingale differences, the obtained rates are always faster than those obtained in case of usual partial sums.

math.PR↗
arXiv2024

Rates in the central limit theorem for random projections of Martingales

In this paper, we consider partial sums of martingale differences weighted by random variables drawn uniformly on the sphere, and globally independent of the martingale differences. Combining Lindeberg's method and a series of arguments due to Bobkov, Chistyakov and G{\"o}tze, we show that the Kolmogorov distance between the distribution of these weighted sums and the limiting Gaussian is "super-fast" of order (log n)^2 /n, under conditions allowing us to control the higher-order conditional moments of the martingale differences. We give an application of this result to the least squares estimator of the slope in the linear model with Gaussian design.

math.PR↗
arXiv2024

Asymptotic confidence interval for R2 in multiple linear regression

Following White's approach of robust multiple linear regression, we give asymptotic confidence intervals for the multiple correlation coefficient R2 under minimal moment conditions. We also give the asymptotic joint distribution of the empirical estimators of the individual R2's. Through different sets of simulations, we show that the procedure is indeed robust (contrary to the procedure involving the near exact distribution of the empirical estimator of R2 is the multivariate Gaussian case) and can be also applied to count linear regression.

math.ST↗
arXiv2023

Rates in almost sure invariance principle for nonuniformly hyperbolic maps

We prove the Almost Sure Invariance Principle (ASIP) with close to optimal error rates for nonuniformly hyperbolic maps. We do not assume exponential contraction along stable leaves, therefore our result covers in particular slowly mixing invertible dynamical systems as Bunimovich flowers, billiards with flat points as in Chernov and Zhang (2005) and Wojtkowski' (1990) system of two falling balls. For these examples, the ASIP is a new result, not covered by prior works for various reasons, notably because in absence of exponential contraction along stable leaves, it is challenging to employ the so-called Sinai's trick (Sinai 1972, Bowen 1975) of reducing a nonuniformly hyperbolic system to a nonuniformly expanding one. Our strategy follows our previous papers on the ASIP for nonuniformly expanding maps, where we build a semiconjugacy to a specific renewal Markov shift and adapt the argument of Berkes, Liu and Wu (2014). The main difference is that now the Markov shift is two-sided, the observables depend on the full trajectory, both the future and the past.

math.DS↗
arXiv2023

Deviation inequalities for dependent sequences with applications to strong approximations

In this paper, we give precise rates of convergence in the strong invariance principle for stationary sequences of bounded real-valued random variables satisfying weak dependence conditions. One of the main ingredients is a new Fuk-Nagaev type inequality for a class of weakly dependent sequences. We describe also several classes of processes to which our results apply.

math.PR↗
arXiv2023

Limit theorems for iid products of positive matrices

We study stochastic properties of the norm cocycle associated with iid products of positive matrices. We obtain the almost sure invariance principle (ASIP) with rate o(n 1/p) under the optimal condition of a moment or order p > 2 and the Berry-Esseen theorem with rate O(1/ $\sqrt$ n) under the optimal condition of a moment of order 3. The results are also valid for the matrix norm. For the matrix coefficients, we also have the ASIP but we obtain only partial results for the Berry-Esseen theorem. The proofs make use of coupling coefficients that surprisingly decay exponentially fast to 0 while there is only a polynomial decay in the case of invertible matrices. All the results are actually valid in the context of iid products of matrices leaving invariant a suitable cone.

math.PR↗
arXiv2022

Berry-Esseen type bounds for the Left Random Walk on GL d (R) under polynomial moment conditions

Let $A_n= \varepsilon_n \cdots \varepsilon_1$, where $(\varepsilon_n)_{n \geq 1}$ is a sequence of independent random matrices taking values in $ GL_d(\mathbb R)$, $d \geq 2$, with common distribution $\mu$. In this paper, under standard assumptions on $\mu$ (strong irreducibility and proximality), we prove Berry-Esseen type theorems for $\log ( \Vert A_n \Vert)$ when $\mu$ has a polynomial moment. More precisely, we get the rate $((\log n) / n)^{q/2-1}$ when $\mu$ has a moment of order $q \in ]2,3]$ and the rate $1/ \sqrt{n} $ when $\mu$ has a moment of order $4$, which significantly improves earlier results in this setting.

math.PR↗
arXiv2018

Rates in almost sure invariance principle for quickly mixing dynamical systems

For a large class of quickly mixing dynamical systems, we prove that the error in the almost sure approximation with a Brownian motion is of order O((log n)^a) with a $\ge$ 2. Specifically, we consider nonuniformly expanding maps with exponential and stretched exponential decay of correlations, with one-dimensional H{\"o}lder continuous observables.

math.PR↗
arXiv2016

Large and moderate deviations for bounded functions of slowly mixing markov chains

We consider Markov chains which are polynomially mixing, in a weak sense expressed in terms of the space of functions on which the mixing speed is controlled. In this context, we prove polynomial large and moderate deviations inequalities. These inequalities can be applied in various natural situations coming from probability theory or dynamical systems. Finally, we discuss examples from these various settings showing that our inequalities are sharp.

math.PR↗
arXiv2016

A deviation bound for $\alpha$-dependent sequences with applications to intermittent maps

We prove a deviation bound for the maximum of partial sums of functions of $\alpha$-dependent sequences as defined in Dedecker, Gou{\"e}zel and Merlev{\`e}de (2010). As a consequence, we extend the Rosenthal inequality of Rio (2000) for $\alpha$-mixing sequences in the sense of Rosenblatt (1956) to the larger class of $\alpha$-dependent sequences. Starting from the deviation inequality, we obtain upper bounds for large deviations and an H{\"o}lderian invariance principle for the Donsker line. We illustrate our results through the example of intermittent maps of the interval, which are not $\alpha$-mixing in the sense of Rosenblatt.

math.PR↗