Pregled bibliografske jedinice broj: 451637
Bayesian Networks Structure Design Procedure for Medical Decision Support
Bayesian Networks Structure Design Procedure for Medical Decision Support // 74th Annual Meeting of Mississippi Academy of Sciences / NA (ur.).
Hattiesburg (MS), 2010. str. 121-121 (predavanje, međunarodna recenzija, sažetak, znanstveni)
CROSBI ID: 451637 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Bayesian Networks Structure Design Procedure for Medical Decision Support
Autori
Gadewadikar, Jyotirmay ; Kuljaca, Ognjen ; Agyepong, Kwabena ;
Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni
Izvornik
74th Annual Meeting of Mississippi Academy of Sciences
/ NA - Hattiesburg (MS), 2010, 121-121
Skup
74th Annual Meeting of Mississippi Academy of Sciences
Mjesto i datum
Hattiesburg (MS), Sjedinjene Američke Države, 02.2010
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
Bayesian network; design procedure; breast cancer; decision support
Sažetak
A Bayesian Network is a graphical model that represents relationships of probabilistic nature among variables of interest. This work presents Bayesian network Structure Design Procedure introduction towards implementation of an automated breast cancer detection support tool. The work presented extends, elaborates, and concocts wealth of knowledge with available open source tools, and the documentation available in the popular literature. The literature on efficient construction of Bayesian networks for complex domains is very limited. Moreover, it often overlooks the fact that in practice it is very desirable to be able to deliver a diagnostic tool of limited performance early in the design process and refine it progressively as more information becomes available. The available literature can be selected as a guide for construction of Bayesian Networks for Diagnostics, over many available, however, since most of the available work deal with engineering systems diagnostics a fair amount of modifications needs to be done in order to make a Medical Diagnostics Systems. There is a wealth of knowledge in pure engineering diagnostic Bayesian Network domain because of the fact that it is more acceptable to automate a decision process in an engineering system as opposed to a system involving medical diagnostics. The key problem is to construct the model so that we capture all of the important aspects of system reality from the point of view of the diagnostic process. In this work examination and diagnostic process point of view is presented along with the Bayesian Network Model Structure Design procedure.
Izvorni jezik
Engleski
Znanstvena područja
Elektrotehnika
POVEZANOST RADA
Projekti:
192-0361557-1564 - Inteligentno upravljanje hidroenergetskim postrojenjima
Ustanove:
Brodarski institut d.o.o.
Profili:
Ognjen Kuljača
(autor)