Topics in mathematical and applied statistics

Author(s):
Pas, S.L. van der (0000-0002-2448-5378)
Published
2017
Publication type
Doctoral Thesis

Abstract

This thesis is composed of papers on four topics: Bayesian theory for the sparse normal means problem, specifically for the horseshoe prior (Chapters 1-3), Bayesian theory for community detection (Chapter 4), nested model selection (Chapter 5), and the application of competing risk methods in the presence of time-dependent clustering (Chapter 6).

Horseshoe priorFrequentist BayesCommunity detectionHypothesis testingSurvival analysis
Defense date
2017-02-28
Language
English
Supervisor(s)
Vaart, A.W. van der
Committee member(s)
Castillo, I.; Gill, R.D.; Grünwald, P.D.; Gugushvili, S.; Rocková, V.
Qualification
Ph.D.
Degree granting institution
Mathematisch Instituut, Science, Leiden University, Leiden
hdl
1887/46454