Nuclear mass predictions based on a deep neural network and finite-range droplet model (2012)

A neural network with two hidden layers is developed for nuclear mass prediction, based on the finite-range droplet model (FRDM12). Different hyperparameters, including the number of hidden units, choice of activation functions, initializers, and learning rates, are adjusted explicitly and systemati...

Full description

Saved in:
Bibliographic Details
Published in:Chinese physics C 2024-02, Vol.48 (2), p.24102
Main Authors: Yiu 姚, To Chung 道驄, Liang 梁, Haozhao 豪兆, Lee 李, Jenny 曉菁
Format: Article
Language:English
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by cdi_FETCH-LOGICAL-c285t-36330f8f5138ed621173710acaf95d80d351dc94613589118c8d0142e508aa9a3
cites cdi_FETCH-LOGICAL-c285t-36330f8f5138ed621173710acaf95d80d351dc94613589118c8d0142e508aa9a3
container_end_page
container_issue 2
container_start_page 24102
container_title Chinese physics C
container_volume 48
creator Yiu 姚, To Chung 道驄
Liang 梁, Haozhao 豪兆
Lee 李, Jenny 曉菁
description A neural network with two hidden layers is developed for nuclear mass prediction, based on the finite-range droplet model (FRDM12). Different hyperparameters, including the number of hidden units, choice of activation functions, initializers, and learning rates, are adjusted explicitly and systematically. The resulting mass predictions are achieved by averaging the predictions given by several different sets of hyperparameters with different regularizers and seed numbers. This can provide not only the average values of mass predictions but also reliable estimations in the mass prediction uncertainties. The overall root-mean-square deviations of nuclear mass are reduced from 0.603 MeV for the FRDM12 model to 0.200 MeV and 0.232 MeV for the training and validation sets, respectively.
doi_str_mv 10.1088/1674-1137/ad021c
format article
fullrecord <record><control><sourceid>crossref</sourceid><recordid>TN_cdi_crossref_primary_10_1088_1674_1137_ad021c</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>10_1088_1674_1137_ad021c</sourcerecordid><originalsourceid>FETCH-LOGICAL-c285t-36330f8f5138ed621173710acaf95d80d351dc94613589118c8d0142e508aa9a3</originalsourceid><addsrcrecordid>eNo9kE1LxDAURYMoWEf3LrPURZ33kqZNlzLoKAy60aWUZ_Iq1X6RdBD_vS0jru7ici-cI8Qlwg2CtWvMiyxF1MWaPCh0RyJRYGyao1bHIvmvT8VZjJ8AeTavEvH2tHctU5AdxSjHwL5xUzP0Ub5TZC-HXpL0zKPseR-onWP6HsKXpN7LuumbidNA_QdLH4ax5Ul2g-dWXilAdX0uTmpqI1_85Uq83t-9bB7S3fP2cXO7S52yZkp1rjXUtjaoLftcIRa6QCBHdWm8Ba8NeldmM4uxJaJ11gNmig1YopL0SsDh14UhxsB1NYamo_BTIVSLnmrhrxb-6qBH_wIlH1c1</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Nuclear mass predictions based on a deep neural network and finite-range droplet model (2012)</title><source>Institute of Physics:Jisc Collections:IOP Publishing Read and Publish 2024-2025 (Reading List)</source><creator>Yiu 姚, To Chung 道驄 ; Liang 梁, Haozhao 豪兆 ; Lee 李, Jenny 曉菁</creator><creatorcontrib>Yiu 姚, To Chung 道驄 ; Liang 梁, Haozhao 豪兆 ; Lee 李, Jenny 曉菁</creatorcontrib><description>A neural network with two hidden layers is developed for nuclear mass prediction, based on the finite-range droplet model (FRDM12). Different hyperparameters, including the number of hidden units, choice of activation functions, initializers, and learning rates, are adjusted explicitly and systematically. The resulting mass predictions are achieved by averaging the predictions given by several different sets of hyperparameters with different regularizers and seed numbers. This can provide not only the average values of mass predictions but also reliable estimations in the mass prediction uncertainties. The overall root-mean-square deviations of nuclear mass are reduced from 0.603 MeV for the FRDM12 model to 0.200 MeV and 0.232 MeV for the training and validation sets, respectively.</description><identifier>ISSN: 1674-1137</identifier><identifier>EISSN: 2058-6132</identifier><identifier>DOI: 10.1088/1674-1137/ad021c</identifier><language>eng</language><ispartof>Chinese physics C, 2024-02, Vol.48 (2), p.24102</ispartof><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c285t-36330f8f5138ed621173710acaf95d80d351dc94613589118c8d0142e508aa9a3</citedby><cites>FETCH-LOGICAL-c285t-36330f8f5138ed621173710acaf95d80d351dc94613589118c8d0142e508aa9a3</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>314,780,784,27924,27925</link.rule.ids></links><search><creatorcontrib>Yiu 姚, To Chung 道驄</creatorcontrib><creatorcontrib>Liang 梁, Haozhao 豪兆</creatorcontrib><creatorcontrib>Lee 李, Jenny 曉菁</creatorcontrib><title>Nuclear mass predictions based on a deep neural network and finite-range droplet model (2012)</title><title>Chinese physics C</title><description>A neural network with two hidden layers is developed for nuclear mass prediction, based on the finite-range droplet model (FRDM12). Different hyperparameters, including the number of hidden units, choice of activation functions, initializers, and learning rates, are adjusted explicitly and systematically. The resulting mass predictions are achieved by averaging the predictions given by several different sets of hyperparameters with different regularizers and seed numbers. This can provide not only the average values of mass predictions but also reliable estimations in the mass prediction uncertainties. The overall root-mean-square deviations of nuclear mass are reduced from 0.603 MeV for the FRDM12 model to 0.200 MeV and 0.232 MeV for the training and validation sets, respectively.</description><issn>1674-1137</issn><issn>2058-6132</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><recordid>eNo9kE1LxDAURYMoWEf3LrPURZ33kqZNlzLoKAy60aWUZ_Iq1X6RdBD_vS0jru7ici-cI8Qlwg2CtWvMiyxF1MWaPCh0RyJRYGyao1bHIvmvT8VZjJ8AeTavEvH2tHctU5AdxSjHwL5xUzP0Ub5TZC-HXpL0zKPseR-onWP6HsKXpN7LuumbidNA_QdLH4ax5Ul2g-dWXilAdX0uTmpqI1_85Uq83t-9bB7S3fP2cXO7S52yZkp1rjXUtjaoLftcIRa6QCBHdWm8Ba8NeldmM4uxJaJ11gNmig1YopL0SsDh14UhxsB1NYamo_BTIVSLnmrhrxb-6qBH_wIlH1c1</recordid><startdate>20240201</startdate><enddate>20240201</enddate><creator>Yiu 姚, To Chung 道驄</creator><creator>Liang 梁, Haozhao 豪兆</creator><creator>Lee 李, Jenny 曉菁</creator><scope>AAYXX</scope><scope>CITATION</scope></search><sort><creationdate>20240201</creationdate><title>Nuclear mass predictions based on a deep neural network and finite-range droplet model (2012)</title><author>Yiu 姚, To Chung 道驄 ; Liang 梁, Haozhao 豪兆 ; Lee 李, Jenny 曉菁</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c285t-36330f8f5138ed621173710acaf95d80d351dc94613589118c8d0142e508aa9a3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Yiu 姚, To Chung 道驄</creatorcontrib><creatorcontrib>Liang 梁, Haozhao 豪兆</creatorcontrib><creatorcontrib>Lee 李, Jenny 曉菁</creatorcontrib><collection>CrossRef</collection><jtitle>Chinese physics C</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Yiu 姚, To Chung 道驄</au><au>Liang 梁, Haozhao 豪兆</au><au>Lee 李, Jenny 曉菁</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Nuclear mass predictions based on a deep neural network and finite-range droplet model (2012)</atitle><jtitle>Chinese physics C</jtitle><date>2024-02-01</date><risdate>2024</risdate><volume>48</volume><issue>2</issue><spage>24102</spage><pages>24102-</pages><issn>1674-1137</issn><eissn>2058-6132</eissn><abstract>A neural network with two hidden layers is developed for nuclear mass prediction, based on the finite-range droplet model (FRDM12). Different hyperparameters, including the number of hidden units, choice of activation functions, initializers, and learning rates, are adjusted explicitly and systematically. The resulting mass predictions are achieved by averaging the predictions given by several different sets of hyperparameters with different regularizers and seed numbers. This can provide not only the average values of mass predictions but also reliable estimations in the mass prediction uncertainties. The overall root-mean-square deviations of nuclear mass are reduced from 0.603 MeV for the FRDM12 model to 0.200 MeV and 0.232 MeV for the training and validation sets, respectively.</abstract><doi>10.1088/1674-1137/ad021c</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 1674-1137
ispartof Chinese physics C, 2024-02, Vol.48 (2), p.24102
issn 1674-1137
2058-6132
language eng
recordid cdi_crossref_primary_10_1088_1674_1137_ad021c
source Institute of Physics:Jisc Collections:IOP Publishing Read and Publish 2024-2025 (Reading List)
title Nuclear mass predictions based on a deep neural network and finite-range droplet model (2012)
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-03T18%3A29%3A04IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-crossref&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Nuclear%20mass%20predictions%20based%20on%20a%20deep%20neural%20network%20and%20finite-range%20droplet%20model%20(2012)&rft.jtitle=Chinese%20physics%20C&rft.au=Yiu%20%E5%A7%9A,%20To%20Chung%20%E9%81%93%E9%A9%84&rft.date=2024-02-01&rft.volume=48&rft.issue=2&rft.spage=24102&rft.pages=24102-&rft.issn=1674-1137&rft.eissn=2058-6132&rft_id=info:doi/10.1088/1674-1137/ad021c&rft_dat=%3Ccrossref%3E10_1088_1674_1137_ad021c%3C/crossref%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c285t-36330f8f5138ed621173710acaf95d80d351dc94613589118c8d0142e508aa9a3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true