Conditional independence testing for variable selection and causal inference
Abstract/Contents
- Abstract
- Conditional independence testing is a central statistical topic, and here we report on recent advances leveraging such tests for variable selection and causal inference. First, we consider the generation of model-X knockoffs, developing an efficient sampler for arbitrary graphical models as well as proving a lower bound on the computational complexity of knockoff sampling. Our new knockoff sampler enables the deployment of the principled model-X knockoff framework for variable selection for a wider range of problems. Second, we turn to variable selection with compositional data, developing a penalized fitting procedure and a conditional inference procedure to asses goodness-of-fit. We also briefly discuss the connections between model-X knockoff testing and conditional inference. Lastly, we consider the selection of causal variables in genetics. Using data from parent-offspring trios, we formulate inheritance as a high-dimensional randomized experiment, explain how this corrects for confounders in genome-wide association studies, and develop a set of conditional independence tests compatible with multiple testing procedures to check for causal variants in multiple regions of the genome
Description
Type of resource | text |
---|---|
Form | electronic resource; remote; computer; online resource |
Extent | 1 online resource |
Place | California |
Place | [Stanford, California] |
Publisher | [Stanford University] |
Copyright date | 2020; ©2020 |
Publication date | 2020; 2020 |
Issuance | monographic |
Language | English |
Creators/Contributors
Author | Bates, Stephen Douglas |
---|---|
Degree supervisor | Candès, Emmanuel J. (Emmanuel Jean) |
Thesis advisor | Candès, Emmanuel J. (Emmanuel Jean) |
Thesis advisor | Tibshirani, Robert |
Thesis advisor | Wager, Stefan |
Degree committee member | Tibshirani, Robert |
Degree committee member | Wager, Stefan |
Associated with | Stanford University, Department of Statistics |
Subjects
Genre | Theses |
---|---|
Genre | Text |
Bibliographic information
Statement of responsibility | Stephen Douglas Bates |
---|---|
Note | Submitted to the Department of Statistics |
Thesis | Thesis Ph.D. Stanford University 2020 |
Location | electronic resource |
Access conditions
- Copyright
- © 2020 by Stephen Douglas Bates
Also listed in
Loading usage metrics...