Estimation of integrated squared density derivatives from a contaminated sample

By Aurore Delaigle, Irène Gijbels. 2002

86 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/ff50d98f3de2e181c2307077c3486146cdda3266

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistics and Probability

Citations in the corpus

Density Estimation in the Presence of Heteroscedastic Measurement Error (2008, methodology): “Other recent progress in this area has been made by Delaigle (2007a), Delaigle, Hall, and Qiu (2006), Delaigle and Gijbels (2002, 2004), Meister (2004, 2006), and Schennach (2004). Stefanski and Bay (1996) propose an alternative non-parametric approach where the effect of measurement error (assuming”
Bootstrap bandwidth selection in kernel density estimation from a contaminated sample (2004, methodology): “Such a bandwidth has been proposed by Delaigle and Gijbels (2002) , and is easy to use in practice.”
A ridge-parameter approach to deconvolution (2007, result): “More recent work, which also surveys earlier research, includes that of van Es, Spreij and van Zanten [29], Delaigle and Gijbels [[8, 9, 10]], Meister [[23, 24]] and Comte, Rozenholc and Taupin [[6, 7]].”
Methodology and convergence rates for functional linear regression (2007, background): “The estimation of slope and intercept parameters in functional linear regression has points in common with a range of smoothing and deconvolution problems where dimension reduction is involved; see, for example, [9, 12, 13, 28].”
On deconvolution with repeated measurements (2008, methodology): “In the density case, we suggest using ĥPI, the plug-in bandwidth of Delaigle and Gijbels (2002, 2004), where the characteristic function of the error is replaced by (2.3).”
Methodology for non‐parametric deconvolution when the error distribution is unknown (2016, methodology): “…FU , implemented as in Section 4.1, the standard deconvolution estimator f̂ DEC at equation (4.1) for known FU with the two-stage plug-in bandwidth of Delaigle and Gijbels (2002, 2004), and the naive estimator f̂ naive (the standard kernel estimator of fW computed from the Wis) with the Sheather…”
Non‐parametric confidence bands in deconvolution density estimation (2007, methodology): “…is the density f of the Zi, which is related to the density g of the Xi via g=f Åψ, .2/ the convolution of f and ψ. Recovering f from the noisy observations X1, . . . , Xn is therefore called the deconvolution problem (e.g. Fan (1991a, b), Diggle and Hall (1993) and Delaigle and Gijbels (2002)).”