Naslov
Learning Support Vector Regression Models for Fast Radiation Dose Rate Calculations
(Learning Support Vector Regression Models for Fast Radiation Dose Rate)

Autori
Trontl, Krešimir ; Šmuc, Tomislav ; Pevec, Dubravko

Vrsta, podvrsta i kategorija rada
Poglavlja u knjigama, znanstveni

Knjiga
Machine Learning Research Progress

Urednik/ci
Peters, Hannah ; Vogel, Mia

Izdavač
Nova Science Publishers

Grad
New York (NY)

Godina
2010

Raspon stranica
427-462

ISBN
978-1-60456-646-8

Ključne riječi
gamma dose rate, multi-layer shields, support vector regression, buildup factor

Sažetak
In this chapter we consider the application of Support Vector Regression (SVR) in the field of radiation dose rate calculations, namely determination of gamma ray dose buildup factors. We demonstrate that SVR model for buildup factor determination can be applied as a fast engineering tool, replacing more traditional approaches based on semi-empirical formulas. More important is the fact that using general regression model like SVR in conjunction with machine learning methodology for the development and evaluation of learned models, provides a general approach for replacing complex simulation models. Therefore, we attempt to summarize research activities in a set of guidelines and procedures for performing an optimized search for the SVR model, for similar types of physical problems.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika

POVEZANOST RADA