dorsal/arxiv
View SchemaApplication Of Support Vector Machines To Global Prediction Of Nuclear Properties
| Authors | John W. Clark, Haochen Li |
|---|---|
| Categories | |
| ArXiv ID | nucl-th/0603037 |
| URL | https://arxiv.org/abs/nucl-th/0603037 |
| DOI | 10.1142/S0217979206036053 |
| Journal | Int.J.Mod.Phys.B20:5015-5029,2006 |
Abstract
Advances in statistical learning theory present the opportunity to develop statistical models of quantum many-body systems exhibiting remarkable predictive power. The potential of such ``theory-thin'' approaches is illustrated with the application of Support Vector Machines (SVMs) to global prediction of nuclear properties as functions of proton and neutron numbers $Z$ and $N$ across the nuclidic chart. Based on the principle of structural-risk minimization, SVMs learn from examples in the existing database of a given property $Y$, automatically and optimally identify a set of ``support vectors'' corresponding to representative nuclei in the training set, and approximate the mapping $(Z,N) \to Y$ in terms of these nuclei. Results are reported for nuclear masses, beta-decay lifetimes, and spins/parities of nuclear ground states. These results indicate that SVM models can match or even surpass the predictive performance of the best conventional ``theory-thick'' global models based on nuclear phenomenology.
{
"annotation_id": "cb932a5c-af71-4f9e-beac-066b217d559e",
"date_created": "2026-03-02T18:00:08.578000Z",
"date_modified": "2026-03-02T18:00:08.578000Z",
"file_hash": "929c1a0a4f8c3af7b3a509c93e92dbae28b0771098648c37dcb947a1c570fe67",
"private": false,
"record": {
"abstract": "Advances in statistical learning theory present the opportunity to develop\nstatistical models of quantum many-body systems exhibiting remarkable\npredictive power. The potential of such ``theory-thin\u0027\u0027 approaches is\nillustrated with the application of Support Vector Machines (SVMs) to global\nprediction of nuclear properties as functions of proton and neutron numbers $Z$\nand $N$ across the nuclidic chart. Based on the principle of structural-risk\nminimization, SVMs learn from examples in the existing database of a given\nproperty $Y$, automatically and optimally identify a set of ``support vectors\u0027\u0027\ncorresponding to representative nuclei in the training set, and approximate the\nmapping $(Z,N) \\to Y$ in terms of these nuclei. Results are reported for\nnuclear masses, beta-decay lifetimes, and spins/parities of nuclear ground\nstates. These results indicate that SVM models can match or even surpass the\npredictive performance of the best conventional ``theory-thick\u0027\u0027 global models\nbased on nuclear phenomenology.",
"arxiv_id": "nucl-th/0603037",
"authors": [
"John W. Clark",
"Haochen Li"
],
"categories": [
"nucl-th"
],
"doi": "10.1142/S0217979206036053",
"journal_ref": "Int.J.Mod.Phys.B20:5015-5029,2006",
"title": "Application Of Support Vector Machines To Global Prediction Of Nuclear Properties",
"url": "https://arxiv.org/abs/nucl-th/0603037"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "9ebd6909-455c-4d1c-be71-32be148aa6f3",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}