Better predictions when models are wrong or underspecified

Leiden Repository

Better predictions when models are wrong or underspecified

Type: Doctoral Thesis
Title: Better predictions when models are wrong or underspecified
Author: Ommen, M. van
Issue Date: 2015-06-10
Keywords: Inconsistency
Probability updating
Monty Hall
Bayesian statistics
Abstract: Many statistical methods rely on models of reality in order to learn from data and to make predictions about future data. By necessity, these models usually do not match reality exactly, but are either wrong (none of the hypotheses in the model provides an accurate description of reality) or underspecified (the hypotheses in the model describe only part of the data). In this thesis, we discuss three scenarios involving models that are wrong or underspecified. In each case, we find that standard statistical methods may fail, sometimes dramatically, and present different methods that continue to perform well even if the models are wrong or underspecified. The first two of these scenarios involve regression problems and investigate AIC (Akaike's Information Criterion) and Bayesian statistics. The third scenario has the famous Monty Hall problem as a special case, and considers the question how we can update our belief about an unknown outcome given new evidence when the precise relation between outcome and evidence is unknown.
Promotor: Supervisor: Grünwald P.D.
Faculty: Faculty of Science
University: Leiden University
Uri: urn:isbn:9789462596894

Files in this item

Description Size View
application/pdf Full text 2.691Mb View/Open
application/pdf Cover 29.98Mb View/Open
application/pdf Title page_Contents_ List of figures 262.3Kb View/Open
application/pdf Chapter 1 Introduction 385.0Kb View/Open
application/pdf Chapter 2 421.5Kb View/Open
application/pdf Chapter 3 883.9Kb View/Open
application/pdf Chapter 4 452.8Kb View/Open
application/pdf Chapter 5 748.9Kb View/Open
application/pdf Chapter 6 769.7Kb View/Open
application/pdf Chapter 7 406.0Kb View/Open
application/pdf Chapter 8 423.7Kb View/Open
application/pdf Bibliography_Index 251.5Kb View/Open
application/pdf Summary in Dutch 186.2Kb View/Open
application/pdf Acknowledgements_Curriculum Vitae 133.3Kb View/Open
application/pdf Propositions 135.2Kb View/Open

This item appears in the following Collection(s)