Structure learning in high-dimensional graphical models

In this thesis, we develop efficient and provably consistent algorithms for learning the structure of undirected and directed (causal) graphical models in the high-dimensional setting. Structure learning in graphical models is a central problem in statistics with numerous applications including learning gene regulatory networks from RNA-seq data and learning the dependence structure among stocks from financial time series. Part I of this thesis investigates the problem of learning causal directed acyclic graph (DAG) models from a combination of observational and interventional data. While previous methods considered greedy search algorithms on the space of graphs, we propose to view a DAG as given by a permutation and an undirected graph and instead consider greedy search on the smaller space of permutations. We present the first consistency guarantees of a permutation-based greedy search algorithm based on observational data.

Name thesis:degree_name

Doctoral

Department dc:contributor.department

Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2019

Author dc:creator

• Wang, Yuhao,Ph.D.Massachusetts Institute of Technology.

Advisor dc:contributor.advisor

• Caroline Uhler.

• Electrical Engineering and Computer Science.

Statement dc:rights

• MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission.

Licence dc:rights.uri

• http://dspace.mit.edu/handle/1721.1/7582

Language dc:language.iso

eng

Handle dc:identifier.uri

https://hdl.handle.net/1721.1/122688

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/122688