Real-time gamma-neutron discrimination with a trainable
polynomial kernel

D 2025

Real-time gamma-neutron discrimination with a trainable polynomial kernel

HLAVINKA, Matyáš; Zdeněk MATĚJ; Jiří ČULEN; Jan KRÁL; Václav PŘENOSIL et al.

Základní údaje

Originální název

Real-time gamma-neutron discrimination with a trainable polynomial kernel

Autoři

HLAVINKA, Matyáš; Zdeněk MATĚJ ; Jiří ČULEN; Jan KRÁL a Václav PŘENOSIL

Vydání

Les Ulis, EPJ Web Conf. Volume 338, 2025 ANIMMA 2025 – Advancements in Nuclear Instrumentation Measurement Methods and their Applications, od s. 1-7, 7 s. 2025

Nakladatel

EDP Sciences

Další údaje

Jazyk

angličtina

Typ výsledku

Stať ve sborníku

Obor

10201 Computer sciences, information science, bioinformatics

Utajení

není předmětem státního či obchodního tajemství

Forma vydání

elektronická verze "online"

Odkazy

URL

Označené pro přenos do RIV

Ano

Kód RIV

RIV/00216224:14330/25:00142750

Organizační jednotka

Fakulta informatiky

ISSN

DOI

https://doi.org/10.1051/epjconf/202533810004

EID Scopus

2-s2.0-105024528968

Klíčová slova anglicky

gamma-neutron discrimination; support vector machines; field-programmable gate array

Změněno: 9. 4. 2026 09:53, RNDr. Pavel Šmerk, Ph.D.

Anotace

V originále

This paper presents an implementation of gamma-neutron pulse shape discrimination by a support vector machine polynomial decision function in a field-programmable gate array. The training is carried out on a conventional computer using widespread Python libraries. The hardware architecture is designed to allow parameter changes on demand, enabling tuning of hyperparameters and kernel coefficients without requiring re-synthesis. A cubic kernel is compared against a linear kernel which was developed alongside it for non-biased comparison. Both are designed to be viable for real-time classification. The particularities of the designs are explored. The cubic kernel makes use of two stand-alone state machines to keep the sequential data pipelined without interference between the sampled pulses. The results show the trade-off between separation quality, numerical accuracy and physical on-board requirements of the implementations. The separation quality is demonstrated on two datasets, one with a noticeable overlap, to assess any benefits the cubic kernel may bring.

Návaznosti

MUNI/A/1709/2024, interní kód MU

Název: Aplikovaný výzkum na FI: Důvěra v dynamických softwarových ekosystémech, kryptografické implementace včetně uživatelských aspektů, kyberbezpečnostní cvičení, fúze dat pro fyzikální sensory a algoritmy plánování v logistice

Investor: Masarykova univerzita, Aplikovaný výzkum na FI: Důvěra v dynamických softwarových ekosystémech, kryptografické implementace včetně uživatelských aspektů, kyberbezpečnostní cvičení, fúze dat pro fyzikální sensory a algoritmy plánování v logistice

Citovat

HLAVINKA, Matyáš; Zdeněk MATĚJ; Jiří ČULEN; Jan KRÁL a Václav PŘENOSIL. Real-time gamma-neutron discrimination with a trainable polynomial kernel. Online. In EPJ Web Conf. Volume 338, 2025 ANIMMA 2025 – Advancements in Nuclear Instrumentation Measurement Methods and their Applications. Les Ulis: EDP Sciences, 2025, s. 1-7. ISSN 2101-6275. Dostupné z: https://doi.org/10.1051/epjconf/202533810004.

@inproceedings{2535962,
   author = {Hlavinka, Matyáš and Matěj, Zdeněk and Čulen, Jiří and Král, Jan and Přenosil, Václav},
   address = {Les Ulis},
   booktitle = {EPJ Web Conf. Volume 338, 2025 ANIMMA 2025 – Advancements in Nuclear Instrumentation Measurement Methods and their Applications},
   doi = {https://doi.org/10.1051/epjconf/202533810004},
   keywords = {gamma-neutron discrimination; support vector machines; field-programmable gate array},
   howpublished = {elektronická verze "online"},
   language = {eng},
   location = {Les Ulis},
   pages = {1-7},
   publisher = {EDP Sciences},
   title = {Real-time gamma-neutron discrimination with a trainable polynomial kernel},
   url = {https://www.epj-conferences.org/articles/epjconf/abs/2025/23/epjconf_animma2025_10004/epjconf_animma2025_10004.html},
   year = {2025}
}

TY  - CONF
ID  - 2535962
AU  - Hlavinka, Matyáš - Matěj, Zdeněk - Čulen, Jiří - Král, Jan - Přenosil, Václav
PY  - 2025
TI  - Real-time gamma-neutron discrimination with a trainable polynomial kernel
PB  - EDP Sciences
CY  - Les Ulis
KW  - gamma-neutron discrimination
KW  - support vector machines
KW  - field-programmable gate array
UR  - https://www.epj-conferences.org/articles/epjconf/abs/2025/23/epjconf_animma2025_10004/epjconf_animma2025_10004.html
N2  - This paper presents an implementation of gamma-neutron pulse shape discrimination by a support vector machine polynomial decision function in a field-programmable gate array. The training is carried out on a conventional computer using widespread Python libraries. The hardware architecture is designed to allow parameter changes on demand, enabling tuning of hyperparameters and kernel coefficients without requiring re-synthesis. A cubic kernel is compared against a linear kernel which was developed alongside it for non-biased comparison. Both are designed to be viable for real-time classification. The particularities of the designs are explored. The cubic kernel makes use of two stand-alone state machines to keep the sequential data pipelined without interference between the sampled pulses. The results show the trade-off between separation quality, numerical accuracy and physical on-board requirements of the implementations. The separation quality is demonstrated on two datasets, one with a noticeable overlap, to assess any benefits the cubic kernel may bring.
ER  -

HLAVINKA, Matyáš; Zdeněk MATĚJ; Jiří ČULEN; Jan KRÁL a Václav PŘENOSIL. Real-time gamma-neutron discrimination with a trainable polynomial kernel. Online. In \textit{EPJ Web Conf. Volume 338, 2025 ANIMMA 2025 – Advancements in Nuclear Instrumentation Measurement Methods and their Applications}. Les Ulis: EDP Sciences, 2025, s.~1-7. ISSN~2101-6275. Dostupné z: https://doi.org/10.1051/epjconf/202533810004.

Přehled o publikaci