| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 1684343 | Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms | 2006 | 9 Pages |
Abstract
A novel approach is presented to extract relevant parameters associated with the energy loss of ejectiles from nuclear reactions obtained by digitizing the signals of a Bragg curve spectrometer. New and more powerful computational paradigms allow a more thorough pulse-shape analysis. This is fulfilled using a back-propagation artificial neural network as a pattern identifier. The known problem of over-training is discussed.
Related Topics
Physical Sciences and Engineering
Materials Science
Surfaces, Coatings and Films
Authors
J.J. Vega, R. Reynoso,