General
Minimax estimation for simultaneous deconvolution model in random design using wavelets
arXiv stat.TH0 views
arXiv:2611.23439v1 Announce Type: new Abstract: Simultaneous deconvolution model is investigated when the design points are random and follow bounded densities and errors are either $i.i.d.$ or long-memory dependent. We propose wavelet hard-thresholding estimators, and show that they achieve asymptotically near-optimal convergence rates for the $L^2$-risk when the response function belongs to Besov space under both $i.i.d.$ errors and homoskedastic, long-memory errors. However, under heteroskedastic, long-memory errors the estimator is not optimal. For the $i.i.d.$ errors, the convergence rates are {identical} to those found in simultaneous deconvolution under regular design {and $i.i.d.$ errors}. However, for long-memory dependent errors, the convergence rates are new {and completely different from those in simultaneous deconvolution under regular design and long-memory errors}.
Source:https://arxiv.org/abs/2611.23439PreviousToward Agentic Retrieval-Augmented Generation for Enterprise Retail Workflows over Heterogeneous Data SourcesNextFederated Differential Privacy with Batch Contamination
Related posts
If nobody writes it down, it did not get decided
0dev.to Productivity
Micrograph B-factor estimation via Wilson statistics
0arXiv eess.IV
Greedy Vector Approximation Based on Rotation for Multiplierless Complex Products
0arXiv eess.SP
Toward Massive MIMO HAPS Networks: Antenna Architectures and Design
0arXiv eess.SP
Deep-Learning-Assisted Full Waveform Inversion for Attenuation Reconstruction in Ultrasound Computed Tomography
0arXiv eess.SP
Two is better than one: optimizing binary rating scales for online platforms
0arXiv cs.MA