User profiles for N. Ignatiadis
Nikolaos IgnatiadisStatistics & Data Science, University of Chicago Verified email at uchicago.edu Cited by 1001 |
Data-driven hypothesis weighting increases detection power in genome-scale multiple testing
Hypothesis weighting improves the power of large-scale multiple testing. We describe
independent hypothesis weighting (IHW), a method that assigns weights using covariates …
independent hypothesis weighting (IHW), a method that assigns weights using covariates …
[HTML][HTML] Mapping lung cancer epithelial-mesenchymal transition states and trajectories with single-cell resolution
… downsampled state (E1 cells n = 2394, E2 cells n = 3284, E3 cells n = 1316, pEMT1 cells …
transition distributions (bootstrap analysis based on n = 10,000 cells for each of seven …
transition distributions (bootstrap analysis based on n = 10,000 cells for each of seven …
Covariate powered cross-weighted multiple testing
N Ignatiadis, W Huber - Journal of the Royal Statistical Society …, 2021 - academic.oup.com
… Hence we build upon the hypothesis-splitting idea of Ignatiadis et al. (2016) and demonstrate
that it can be used not merely as a heuristic, but instead as a theoretically grounded and …
that it can be used not merely as a heuristic, but instead as a theoretically grounded and …
Net benefit, calibration, threshold selection, and training objectives for algorithmic fairness in healthcare
A growing body of work uses the paradigm of algorithmic fairness to frame the development
of techniques to anticipate and proactively mitigate the introduction or exacerbation of health …
of techniques to anticipate and proactively mitigate the introduction or exacerbation of health …
E-values as unnormalized weights in multiple testing
We study how to combine p-values and e-values, and design multiple testing procedures
where both p-values and e-values are available for every hypothesis. Our results provide a …
where both p-values and e-values are available for every hypothesis. Our results provide a …
Creating functional engineered variants of the single-module non-ribosomal peptide synthetase IndC by T domain exchange
… The number of data points for calibration was N = 714. (c) A boxplot of the distribution of
best estimates for parameter values obtained for the four IndC variants. (d) A surface plot color-…
best estimates for parameter values obtained for the four IndC variants. (d) A surface plot color-…
Noise-induced randomization in regression discontinuity designs
… Further suppose that the sequence of weighting kernels γ(n) … This dataset has scaled
mathematics test scores for n = 18, 174 children from kindergarten to fifth grade. Furthermore, each …
mathematics test scores for n = 18, 174 children from kindergarten to fifth grade. Furthermore, each …
Treatment heterogeneity with survival outcomes
This chapter is accompanied by survlearners, a package that provides well-documented
implementations of the conditional average treatment effects (CATE) estimation strategies …
implementations of the conditional average treatment effects (CATE) estimation strategies …
Covariate-powered empirical Bayes estimation
N Ignatiadis, S Wager - Advances in Neural Information …, 2019 - proceedings.neurips.cc
We study methods for simultaneous analysis of many noisy experiments in the presence of
rich covariate information. The goal of the analyst is to optimally estimate the true effect …
rich covariate information. The goal of the analyst is to optimally estimate the true effect …
Confidence intervals for nonparametric empirical Bayes analysis
N Ignatiadis, S Wager - Journal of the American Statistical …, 2022 - Taylor & Francis
… In Section 2 we explain how for common choices of G and F n ( α ) , (7) can be computed by
… n = F n ( α n ) with α n → 0 and a pilot estimate f n ( · ) of the marginal density f G ( · ) (5). F n …
… n = F n ( α n ) with α n → 0 and a pilot estimate f n ( · ) of the marginal density f G ( · ) (5). F n …