Nonparametric Inference under B-bits Quantization
arXiv:1901.08571
Abstract
Statistical inference based on lossy or incomplete samples is often needed in research areas such as signal/image processing, medical image storage, remote sensing, signal transmission. In this paper, we propose a nonparametric testing procedure based on samples quantized to bits through a computationally efficient algorithm. Under mild technical conditions, we establish the asymptotic properties of the proposed test statistic and investigate how the testing power changes as increases. In particular, we show that if exceeds a certain threshold, the proposed nonparametric testing procedure achieves the classical minimax rate of testing (Shang and Cheng, 2015) for spline models. We further extend our theoretical investigations to a nonparametric linearity test and an adaptive nonparametric test, expanding the applicability of the proposed methods. Extensive simulation studies {together with a real-data analysis} are used to demonstrate the validity and effectiveness of the proposed tests.
References in corpus (7)
- Multivariate normal approximation with Stein's method of exchangeable pairs under a general linearity condition
- Nonparametric inference in generalized functional linear models
- Local and global asymptotic inference in smoothing spline models
- Joint asymptotics for semi-nonparametric regression models with partially linear structure
- Nonparametric Testing under Random Projection
- Quantized Estimation of Gaussian Sequence Models in Euclidean Balls
- Quantized Nonparametric Estimation over Sobolev Ellipsoids