Sparse Density Estimation with Measurement Errors

TL;DR

This paper introduces a novel sparse density estimation method that accounts for measurement errors using a weighted Elastic-net penalization, providing theoretical guarantees and demonstrating superior performance in simulations and meteorology data.

Contribution

It proposes a new measurement error-aware sparse density estimation technique with theoretical oracle inequalities and support recovery guarantees, outperforming conventional methods.

Findings

Significant improvement over traditional approaches in simulations.

Effective support recovery with high probability.

Superior detection of multi-modal densities in meteorology data.

Abstract

This paper aims to build an estimate of an unknown density of the data with measurement error as a linear combination of functions from a dictionary. Inspired by the penalization approach, we propose the weighted Elastic-net penalized minimal $\ell_2$-distance method for sparse coefficients estimation, where the adaptive weights come from sharp concentration inequalities. The optimal weighted tuning parameters are obtained by the first-order conditions holding with a high probability. Under local coherence or minimal eigenvalue assumptions, non-asymptotical oracle inequalities are derived. These theoretical results are transposed to obtain the support recovery with a high probability. Then, some numerical experiments for discrete and continuous distributions confirm the significant improvement obtained by our procedure when compared with other conventional approaches. Finally, the application is performed in a meteorology data set. It shows that our method has potency and superiority of detecting the shape of multi-mode density compared with other conventional approaches.