论文部分内容阅读
提出一种按键分类的原子电性距离矢量(B-AEDV),描述各异香豆精类化合物分子中不同等价碳原子的化学微环境,结合γ效应及原子自身杂化状态指数(AHSI),以多元线性回归建立~(13)C核磁共振定量结构波谱关系模型,复相关系数R_(M1)为0.971,标准误差(SD_(M1))为7.769。经留一法交互检验的复相关系数(R_(CV1))为0.961,标准误差(SD_(CV1))为8.917。随机抽出两部样本进行外部检验,得到预测集的复相关系数(R_(test1)及R_(test2))分别为0.900和0.960,标准误差(SD_(test2)及SD_(test2))分别为5.931和8.861。研究结果表明,使用该方法所建模型具有良好的预测能力和稳定性。
In this paper, we propose a button-classified atomic electrical distance vector (B-AEDV) to describe the chemical microenvironment of different isovalent coumarin compounds with different equivalent carbon atoms. Combining the γ effect with atomic self-hybridization state index (AHSI) (13) C nuclear magnetic resonance quantitative structure spectral model was established by multiple linear regression. The complex correlation coefficient R_ (M1) was 0.971 and the standard error (SD_ (M1)) was 7.769. The complex correlation coefficient (R_ (CV1)) of 0.971 and the standard error (SD_ (CV1)) of 8.917 by the one-way reciprocal test. Two samples were randomly selected for external testing. The correlation coefficients (R test1 and R test2) of the prediction set were 0.900 and 0.960 respectively, and the standard errors (SD test2 and SD test2) were 5.931 and 8.861. The results show that the proposed model has good predictive ability and stability.