高等学校化学学报

高等学校化学学报 ›› 2012, Vol. 33 ›› Issue (12): 2734.doi: 10.7503/cjcu20120231

• 物理化学 • 上一篇    下一篇

最小二乘支持向量机方法用于提高低水平量子化学方法计算吸收能的精度

高婷1, 韩博1, 李辉1, 李鸿志1, 吕英华1, 苏忠民2   

  1. 1. 东北师范大学计算机科学与信息技术学院, 智能信息处理吉林省高校重点实验室, 长春 130117;
    2. 东北师范大学化学学院, 功能材料化学研究所, 长春 130024
  • 收稿日期:2012-03-16 出版日期:2012-12-10 发布日期:2012-11-20
  • 通讯作者: 吕英华,男,博士,教授,博士生导师,主要从事模式识别研究.E-mail:luyh@nenu.edu.cn苏忠民,男,博士,教授,博士生导师,主要从事量子化学研究.E-mail:zmsu@nenu.edu.cn E-mail:luyh@nenu.edu.cn;zmsu@nenu.edu.cn
  • 基金资助:

    国家"九七三"计划项目(批准号: 2009CB623605)、 国家自然科学基金(批准号: 20903020)和吉林省科技发展计划项目(批准号: 20110364, 20100114)资助.

Improving the Accuracy of Low Level Quantum Chemical Calculation for Absorption Energies Based on Least Squares Support Vector Machine

GAO Ting1, HAN Bo1, LI Hui1, LI Hong-Zhi1, LV Ying-Hua1, SU Zhong-Min2   

  1. 1. Key Laboratory of Intelligent Information Processing of Jilin Province, College of Computer Science and Information Technolgy, Northeast Normal University, Changchun 130117, China;
    2. Institute of Functional Material Chemistry, Faculty of Chemistry, Northeast Normal University, Changchun 130024, China
  • Received:2012-03-16 Online:2012-12-10 Published:2012-11-20

PDF (PC)

被引次数

1

摘要：

运用B3LYP/STO-3G和ZINDO两种低水平的量子化学方法计算了160个有机分子的UV-Vis吸收光谱, 然后提取合适的物理参数, 并以实验值为基础, 引入最小二乘支持向量机方法以提高吸收能的计算值精度. 结果表明, 最小二乘支持向量机方法可有效提高量子化学计算精度, 体系的吸收能误差均方根分别从0.95和0.46 eV降低到0.16和0.15 eV. 最小二乘支持向量机校正方法的引入可在较少的机时和计算资源下得到比单一的量子化学计算方法更为稳定和精确的计算结果, 且可在现有计算条件下预测现有计算能力达不到的精度. 因此, 将最小二乘支持向量机方法用于量子化学数据分析, 为化学研究准确、 快捷地预测分子性质提供了一种新的研究手段.

关键词: 有机小分子, 吸收能, 最小二乘支持向量机, 密度泛函理论

Abstract:

Least squares support vector machine(LS-SVM) was introduced to improve the calculation accuracy of low level density functional theory. As a demonstration, this approach combined low level quantum mechanical calculation with LS-SVM correction has been applied to evaluate the absorption energies of 160 organic molecules. After LS-SVM correction, the root mean square(RMS) deviations of the calculated absorption energies reduce from 0.95 and 0.46 eV to 0.16 and 0.15 eV for B3LYP/STO-3G and ZINDO methods, respectively. The LS-SVM correction may be a potential better method to predict high accurate absorption energies comparing with quantum chemical calculations when time and resource are limited or it is hard to achieve high accurate results by either quantum chemical calculation or measurement. This method provides a new means for accurately and quickly predicting of the molecular properties in quantum chemistry study.

Key words: Small organic molecule, Absorption energy, Least squares support vector machine, Density functional theory

中图分类号: 

TrendMD: