William J. Welch

William J. Welch

Auto-generated profile, based on publicly available information

William J. Welch

Department of Statistics University of British Columbia V6T 1Z4 Vancouver BC Canada

Research Focus & Metrics

Academic Career Length

37Years (1990-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 July 2023 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
Xin DingXin Ding
Xin Ding
YWYongwei Wang
Yongwei Wang
ZXZuheng Xu
Zuheng Xu
WWWilliam J. Welch
William J. Welch
ZWZ. Jane Wang
Z. Jane Wang


1 November 2018 in Hypertension
Verified
Journal Article
Open access
LLLingli Li
Lingli Li
ELEn Yin Lai
En Yin Lai
ZLZaiming Luo
Zaiming Luo
GSGlenn Solis
Glenn Solis
MMMargarida Mendonca
Margarida Mendonca
CWChristopher S Wilcox
Christopher S Wilcox
High salt, Ang II (angiotensin II), and reactive oxygen species enhance progression of chronic kidney disease. We tested the hypothesis that a high salt intake generates specific reactive oxygen species to enhance Ang II contractions of afferent arterioles from mice with reduced renal mass (RRM). C57BL/6 mice were subjected to surgical RRM or sham operations and received 6% or 0.4% NaCl salt diet for 3 months. Ang II contractions were measured in perfused afferent arterioles and superoxide (O 2 − ) and hydrogen peroxide (H 2 O 2 ) by fluorescence microscopy. RRM enhanced the afferent arteriolar gene expression for p47 phox and neutrophil oxidase (NOX) 2 and high salt intake in RRM mice enhanced gene expression for angiotensin type 1 receptors, POLDIP2 and NOX4 and reduced catalase. High salt in mice with RRM enhanced arteriolar O 2 − and H 2 O 2 generation and maximal contractions to Ang II (10 –6 mol/L) that were dependent on O 2 − because they were prevented by gene deletion of p47 phox and on H 2 O 2 because they were prevented by transgenic smooth muscle cell expression of catalase (tg CAT-SMC ) and POLDIP2 gene deletion. Three months of tempol normalized arteriolar reactive oxygen species and Ang II contractions. However, arteriolar contractions to lower concentrations of Ang II (10 –8 to 10 –11 mol/L) were paradoxically inhibited by H 2 O 2 and POLDIP2. In conclusion, both O 2 − from p47 phox /NOX2 and H 2 O 2 from NOX4/POLDIP2 enhance maximal arteriolar Ang II contractions from RRM mice during high salt, but H 2 O 2 and NOX4/POLDIP2 reduce the sensitivity to lower concentrations of Ang II by >100-fold. Tempol prevents all of these changes in function.

1 September 1990 in Journal of the American Statistical Association
Verified
Journal Article
Closed access
WWWilliam J. Welch
William J. Welch
A general method for constructing permutation tests for various experimental designs follows from invariance and sufficiency. In this framework, randomization (or rerandomization) tests are just special cases of permutation tests. The methodology extends the applicability of permutation tests: An example demonstrates a test for an interaction effect in a factorial experiment, a problem thought to be unapproachable by permutation or randomization methods. Tests are constructed as follows. The modeling of block and treatment effects closely parallels classical analysis of variance (ANOVA). The ANOVA assumptions of iid normal random errors are replaced, however, by the much weaker assumption of exchangeability of subsets of random errors. Distributional dependences on the unknown nuisance parameters for blocks, covariates, and untested treatments and on the unknown error distribution are eliminated by invariance and sufficiency, respectively. Initial reduction of the data by invariance appears to ex...

12 March 2020 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
Xin DingXin Ding
Xin Ding
ZWZ. Jane Wang
Z. Jane Wang
WWWilliam J. Welch
William J. Welch
Filtering out unrealistic images from trained generative adversarial networks (GANs) has attracted considerable attention recently. Two density ratio based subsampling methods---Discriminator Rejection Sampling (DRS) and Metropolis-Hastings GAN (MH-GAN)---were recently proposed, and their effectiveness in improving GANs was demonstrated on multiple datasets. However, DRS and MH-GAN are based on discriminator based density ratio estimation (DRE) methods, so they may not work well if the discriminator in the trained GAN is far from optimal. Moreover, they do not apply to some GANs (e.g., MMD-GAN). In this paper, we propose a novel Softplus (SP) loss for DRE. Based on it, we develop a sample-based DRE method in a feature space learned by a specially designed and pre-trained ResNet-34 (DRE-F-SP). We derive the rate of convergence of a density ratio model trained under the SP loss. Then, we propose three different density ratio based subsampling methods (DRE-F-SP+RS, DRE-F-SP+MH, and DRE-F-SP+SIR) for GANs based on DRE-F-SP. Our subsampling methods do not rely on the optimality of the discriminator and are suitable for all types of GANs. We empirically show our subsampling approach can substantially outperform DRS and MH-GAN on a synthetic dataset and the CIFAR-10 dataset, using multiple GANs.

21 October 2022 in Expert Systems with Applications
Verified
Journal Article
Closed access
Xin DingXin Ding
Xin Ding
YWYongwei Wang
Yongwei Wang
ZXZuheng Xu
Zuheng Xu
ZWZ. Jane Wang
Z. Jane Wang
WWWilliam J. Welch
William J. Welch