利用生成式机器学习构建化学启发的动态Ansatz。

IF 2.8 2区 化学 Q3 CHEMISTRY, PHYSICAL
The Journal of Physical Chemistry A Pub Date : 2025-07-03 Epub Date: 2025-06-17 DOI:10.1021/acs.jpca.5c02346
Sonaldeep Halder, Kartikey Anand, Rahul Maitra
{"title":"利用生成式机器学习构建化学启发的动态Ansatz。","authors":"Sonaldeep Halder, Kartikey Anand, Rahul Maitra","doi":"10.1021/acs.jpca.5c02346","DOIUrl":null,"url":null,"abstract":"<p><p>Generative machine learning models like the Restricted Boltzmann Machine (RBM) provide a practical approach for ansatz construction within the quantum computing framework. This work introduces a method that efficiently leverages RBM and many-body perturbative measures to build a compact chemistry-inspired ansatz for determining accurate molecular energetics. By training on low-rank determinants derived from an approximate wave function, RBM predicts the key high-rank determinants that dominate the ground-state wave function. A shallow depth ansatz is constructed to explicitly incorporate these dominant determinants after dynamically decomposing them into low-rank components and applying many-body perturbative measures for further screening. The method requires no additional measurements beyond the initial training phase. Moreover, it incorporates Bayesian hyperparameter optimization for the RBM, ensuring efficient performance with minimal training data during its limited usage. This approach facilitates the efficient computation of molecular properties, paving the way for exploring new chemical phenomena with near-term quantum computers.</p>","PeriodicalId":59,"journal":{"name":"The Journal of Physical Chemistry A","volume":" ","pages":"5889-5900"},"PeriodicalIF":2.8000,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Construction of Chemistry-Inspired Dynamic Ansatz Utilizing Generative Machine Learning.\",\"authors\":\"Sonaldeep Halder, Kartikey Anand, Rahul Maitra\",\"doi\":\"10.1021/acs.jpca.5c02346\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><p>Generative machine learning models like the Restricted Boltzmann Machine (RBM) provide a practical approach for ansatz construction within the quantum computing framework. This work introduces a method that efficiently leverages RBM and many-body perturbative measures to build a compact chemistry-inspired ansatz for determining accurate molecular energetics. By training on low-rank determinants derived from an approximate wave function, RBM predicts the key high-rank determinants that dominate the ground-state wave function. A shallow depth ansatz is constructed to explicitly incorporate these dominant determinants after dynamically decomposing them into low-rank components and applying many-body perturbative measures for further screening. The method requires no additional measurements beyond the initial training phase. Moreover, it incorporates Bayesian hyperparameter optimization for the RBM, ensuring efficient performance with minimal training data during its limited usage. This approach facilitates the efficient computation of molecular properties, paving the way for exploring new chemical phenomena with near-term quantum computers.</p>\",\"PeriodicalId\":59,\"journal\":{\"name\":\"The Journal of Physical Chemistry A\",\"volume\":\" \",\"pages\":\"5889-5900\"},\"PeriodicalIF\":2.8000,\"publicationDate\":\"2025-07-03\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"The Journal of Physical Chemistry A\",\"FirstCategoryId\":\"1\",\"ListUrlMain\":\"https://doi.org/10.1021/acs.jpca.5c02346\",\"RegionNum\":2,\"RegionCategory\":\"化学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"2025/6/17 0:00:00\",\"PubModel\":\"Epub\",\"JCR\":\"Q3\",\"JCRName\":\"CHEMISTRY, PHYSICAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"The Journal of Physical Chemistry A","FirstCategoryId":"1","ListUrlMain":"https://doi.org/10.1021/acs.jpca.5c02346","RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/6/17 0:00:00","PubModel":"Epub","JCR":"Q3","JCRName":"CHEMISTRY, PHYSICAL","Score":null,"Total":0}
引用次数: 0

摘要

像受限玻尔兹曼机(RBM)这样的生成式机器学习模型为量子计算框架内的ansatz构造提供了一种实用的方法。这项工作介绍了一种方法,有效地利用RBM和多体摄动措施来建立一个紧凑的化学启发的ansatz,以确定准确的分子能量。通过训练来自近似波函数的低秩决定因素,RBM预测了主导基态波函数的关键高秩决定因素。在动态分解成低秩分量并应用多体摄动措施进行进一步筛选后,构建浅深度分析来明确地合并这些主导决定因素。该方法在初始训练阶段之外不需要额外的测量。此外,它还对RBM进行了贝叶斯超参数优化,确保在有限的使用中以最少的训练数据实现高效的性能。这种方法促进了分子性质的有效计算,为近期量子计算机探索新的化学现象铺平了道路。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Construction of Chemistry-Inspired Dynamic Ansatz Utilizing Generative Machine Learning.

Generative machine learning models like the Restricted Boltzmann Machine (RBM) provide a practical approach for ansatz construction within the quantum computing framework. This work introduces a method that efficiently leverages RBM and many-body perturbative measures to build a compact chemistry-inspired ansatz for determining accurate molecular energetics. By training on low-rank determinants derived from an approximate wave function, RBM predicts the key high-rank determinants that dominate the ground-state wave function. A shallow depth ansatz is constructed to explicitly incorporate these dominant determinants after dynamically decomposing them into low-rank components and applying many-body perturbative measures for further screening. The method requires no additional measurements beyond the initial training phase. Moreover, it incorporates Bayesian hyperparameter optimization for the RBM, ensuring efficient performance with minimal training data during its limited usage. This approach facilitates the efficient computation of molecular properties, paving the way for exploring new chemical phenomena with near-term quantum computers.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
The Journal of Physical Chemistry A
The Journal of Physical Chemistry A 化学-物理:原子、分子和化学物理
CiteScore
5.20
自引率
10.30%
发文量
922
审稿时长
1.3 months
期刊介绍: The Journal of Physical Chemistry A is devoted to reporting new and original experimental and theoretical basic research of interest to physical chemists, biophysical chemists, and chemical physicists.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信