National and Kapodistrian University of Athens
Bayesian inversion techniques for stochastic differential equations
Abstract
dc:descriptionThis thesis deals with stochastic differential equations in which we are interested inrecovering the initial condition under the uncertainty which exists mainly because ofthe noise. The problem of identifying or recovering the initial condition proves to bean inverse, most of the times ill-posed problem and thus the recovery is intractable.The solutions of such equations are stochastic processes which can be treated aselements of suitable functional spaces such as Hilbert and Banach spaces. Sincethe solutions are elements of such infinite dimensional spaces, there is the necessityto define probability measures in infinite dimensional spaces. We focus our studyin defining and describing Gaussian probability measures defined in Hilbert spaces.The reason to do so, is that we will face the problem of identifying the initial conidition through by setting a Bayesian framework in stochastic differential equationsand the case of the Gaussian distributions is a very common idea applied throughoutthe literature. By establishing a Bayesian method in recovering the initial conditionof an equation, we distinguish from the usual deterministic approaches applied inthese kind of problems and use a probabilistic method instead, which is well knownin finite dimensional problems.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Pasiouras, Alexandros
- Πασιούρας, Αλέξανδρος
Subjects
dc:subject × 12Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/54155