Javascript must be enabled for the correct page display

A new Bayesian network model for learning static and dynamic interactions from temporal data.

Korre, M (2021) A new Bayesian network model for learning static and dynamic interactions from temporal data. Master's Thesis / Essay, Mathematics.

[img]
Preview
Text
mMATH_2021_KorreM.pdf

Download (1MB) | Preview
[img] Text
toestemming.pdf
Restricted to Registered users only

Download (115kB)

Abstract

Bayesian networks are a popular modelling tool for learning the conditional (in-)dependencies among variables. If the data set consists of independent (steady state) observations, (static) Bayesian networks can be applied and the relationships are learned in form of a directed acyclic graph. The edges of the graph represent contemporaneous dependencies. From temporal data (time series) directed graphs can be learned with dynamic Bayesian networks. The edges then represent dynamic dependencies; i.e. dependencies that are subject to a time lag. Within this Master project, static and dynamic Bayesian networks are combined into a new mixed Bayesian network model, which can model contemporaneous and dynamic interactions simultaneously from temporal data. Then, the model is applied for the inference of symptom networks, on data collected from patients in successive clinical stages of psychosis.

Item Type: Thesis (Master's Thesis / Essay)
Supervisor name: Grzegorczyk, M.A.
Degree programme: Mathematics
Thesis type: Master's Thesis / Essay
Language: English
Date Deposited: 16 Mar 2021 14:26
Last Modified: 16 Mar 2021 14:26
URI: https://fse.studenttheses.ub.rug.nl/id/eprint/24067

Actions (login required)

View Item View Item