A Bayesian Solution to the Duhem Problem

Download

A Bayesian Solution to the Duhem Problem by Tunç Uysal.pdf

Date

2022-12

Author

Uysal, Tunç

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

174
views

255
downloads

In his book The Aim and Structure of Physical Theory Pierre Duhem introduces a problem regarding scientific methods which later had been called the Duhem Problem. The Duhem Problem states that when one uses or structures a theory, that theory is not used or structured independently of other theories. Scientific theories are not isolated from each other; we use other theories in the way of constructing or supporting another one. Then, even though a theory is falsified by experiments, we cannot definitely rule out that theory because we do not explicitly know why the theory is falsified. In this thesis, I will examine the viability of Colin Howson and Peter Urbach’s Bayesian Probability focused solution to the Duhem Problem. Bayesian Probability explains the relation between evidence and hypothesis with respect to their probabilities. It can show whether the evidence confirms or disconfirms the hypothesis by determining the probabilities of the hypothesis before and after the evidence is considered. According to Duhem, the Duhem Problem is related to his idea that a crucial experiment, meaning an experiment which definitively shows the superiority of one theory over others, is not possible. This means that when a scientist conducts an experiment, it cannot rule out all possible explanations apart from one because there might always be an explanation of which has not been thought yet. Howson and Urbach claim that individual probabilities of theories can be measured and used to determine where the error lies when theories as a group encounter an error.

Subject Keywords

Pierre Duhem, Bayesian Probability, Howson and Urbach, The Duhem Problem, Bayes’ Theorem

URI

https://hdl.handle.net/11511/101149

Collections

Graduate School of Social Sciences, Thesis

Suggestions

OpenMETU
Core

A comparison of orthogonal cutting data from experiments with three different finite element models Bil, H; Kilic, SE; Tekkaya, AE (Elsevier BV, 2004-07-01) The aim of this study is to compare various simulation models of orthogonal cutting process with each other as well as with the results of various experiments. Commercial implicit finite element codes MSC.Marc, Deform2D and the explicit code Thirdwave AdvantEdge have been used. In simulations, a rigid tool is advanced incrementally into the deformable workpiece which is remeshed whenever needed. In simulations with MSC.Marc and Thirdwave AdvantEdge, there is no separation criterion defined since chip format...
Use of the isolated problem approach for multi-compartment BEM models of electro-magnetic source imaging Gençer, Nevzat Güneri (IOP Publishing, 2005-07-07) The isolated problem approach (IPA) is a method used in the boundary element method (BEM) to overcome numerical inaccuracies caused by the high-conductivity difference in the skull and the brain tissues in the head. Hamalainen and Sarvas (1989 IEEE Trans. Biomed. Eng. 36 165-71) described how the source terms can be updated to overcome these inaccuracies for a three-layer head model. Meijs et al (1989 IEEE Trans. Biomed. Eng. 36 103849) derived the integral equations for the general case where there are an ...
A new outlier detection method based on convex optimization: application to diagnosis of Parkinson's disease TAYLAN, PAKİZE; Yerlikaya-Ozkurt, Fatma; Bilgic Ucak, Burcu; Weber, Gerhard Wilhelm (Informa UK Limited, 2020-12-01) Neuroscience is a combination of different scientific disciplines which investigate the nervous system for understanding of the biological basis. Recently, applications to the diagnosis of neurodegenerative diseases like Parkinson's disease have become very promising by considering different statistical regression models. However, well-known statistical regression models may give misleading results for the diagnosis of the neurodegenerative diseases when experimental data contain outlier observations that l...
A NONLINEAR-PROGRAMMING MODEL FOR PARTIALLY OBSERVABLE MARKOV DECISION-PROCESSES - FINITE-HORIZON CASE Serin, Yaşar Yasemin (1995-11-02) The concept of partially observable Markov decision processes was born to handle the problem of lack of information about the state of a Markov decision process. If the state of the system is unknown to the decision maker then an obvious approach is to gather information that is helpful in selecting an action, This problem was already solved using the theory of Markov processes. We construct a nonlinear programming model for the same problem and develop a solution algorithm that turns out to be a policy ite...
Combining scaffolding for content and scaffolding for dialogue to support conceptual breakthroughs in understanding probability Kazak, Sibel; Wegerif, Rupert; Fujita, Taro (2015-11-01) In this paper, we explore the relationship between scaffolding, dialogue, and conceptual breakthroughs, using data from a design-based research study that focuses on the development of understanding of probability in 10-12 year old students. The aim of the study is to gain insight into how the combination of scaffolding for content using technology and scaffolding for dialogue can facilitate conceptual breakthroughs. We analyse video-recordings and transcripts of pairs and triads of students solving problem...

Citation Formats

T. Uysal, “A Bayesian Solution to the Duhem Problem,” M.A. - Master of Arts, Middle East Technical University, 2022.