Unit:
Τομέας Θεωρητικής ΠληροφορικήςLibrary of the School of Science
Supervisors info:
Ιωαννίδης Ιωάννης
Original Title:
Bayesian δίκτυα και εύρεση δομής με ακριβή αλγόριθμο
Summary:
Structure discovery has been the main focus on multiple research papers during
the last decade. It can be applied in many fields including sports,
entertainment and even medical science. The purpose of this thesis is to
present the basic characteristics of Bayesian networks, to analyse their
structure and to emphasise on the problem of discovering this structure.
Various different solutions are compared and an exact algorithm for Bayesian
structure discovery through data is implemented.
In addition to this, a focal point is the incorporation of this algorithm in
the AITION platform. Finally, experiments are conducted to prove the efficiency
of the algorithm and their results presented thoroughly. .
Keywords:
structure discovery, inference, dynamic programming, exact algorithm, AITION
document.pdf (1 MB)
Open in new window
attachments.zip
4 KB
File access is restricted.