数理医药学杂志
數理醫藥學雜誌
수리의약학잡지
JOURNAL OF MATHEMATICAL MEDICINE
2013年
4期
434-436
,共3页
胡晓娟%李立科%郭东星%王玉贤%曹叔彦
鬍曉娟%李立科%郭東星%王玉賢%曹叔彥
호효연%리립과%곽동성%왕옥현%조숙언
妊娠高血压%主成分%logistic 回归
妊娠高血壓%主成分%logistic 迴歸
임신고혈압%주성분%logistic 회귀
pregnancy-induced hypertension syndrome%principal component%logistic regression analysis
目的:探讨主成分logistic回归分析方法在研究妊娠高血压病人用药与其妊娠结局关系中的作用以及不同药物对妊娠结局的影响,解决多个疾病因素变量存在共线性的问题。方法:对logistic回归模型进行共线性诊断,应用主成分分析方法对存在共线性的变量综合成一个主分量,然后进行logistic回归分析,得到并解释最终的回归模型。结果:不同药物的不同剂量对妊娠结局的影响各不相同。主成分logistic回归在研究妊娠高血压病人药物与妊娠结局关系中有较好的作用。在对疾病因素进行logistic 回归分析时,若变量间存在多重共线性,采用主成分改进的logistic 回归分析能得出更合理的回归模型。
目的:探討主成分logistic迴歸分析方法在研究妊娠高血壓病人用藥與其妊娠結跼關繫中的作用以及不同藥物對妊娠結跼的影響,解決多箇疾病因素變量存在共線性的問題。方法:對logistic迴歸模型進行共線性診斷,應用主成分分析方法對存在共線性的變量綜閤成一箇主分量,然後進行logistic迴歸分析,得到併解釋最終的迴歸模型。結果:不同藥物的不同劑量對妊娠結跼的影響各不相同。主成分logistic迴歸在研究妊娠高血壓病人藥物與妊娠結跼關繫中有較好的作用。在對疾病因素進行logistic 迴歸分析時,若變量間存在多重共線性,採用主成分改進的logistic 迴歸分析能得齣更閤理的迴歸模型。
목적:탐토주성분logistic회귀분석방법재연구임신고혈압병인용약여기임신결국관계중적작용이급불동약물대임신결국적영향,해결다개질병인소변량존재공선성적문제。방법:대logistic회귀모형진행공선성진단,응용주성분분석방법대존재공선성적변량종합성일개주분량,연후진행logistic회귀분석,득도병해석최종적회귀모형。결과:불동약물적불동제량대임신결국적영향각불상동。주성분logistic회귀재연구임신고혈압병인약물여임신결국관계중유교호적작용。재대질병인소진행logistic 회귀분석시,약변량간존재다중공선성,채용주성분개진적logistic 회귀분석능득출경합리적회귀모형。
Objective:To discuss the application of principal component logistic regression analysis meth-ods in researching the relationship between the medications of pregnancy-induced hypertension syndrome and pregnancy outcome .To study the effects of drugs on pregnancy out-come ,and its contribution to solving multivariate collinearity of the effects of drugs .Methods:Firstly ,diagnose the collinearity of logistic regres-sion model ,and then apply the principal component analysis to integrate the variables existing of collinearity into a principal component ,then logistic regression analysis ,access to ,and explain the final regression mod-el .Results:Principal component logistic regression model has great effect on the study of the relationship be-tween drugs for pregnancy-induced hypertension patients and pregnancy outcome ,during the analysis of dis-ease factors logistic regression ,if there was multicollinearity between variables ,the improved use of principal component logistic regression analysis can get to a more reasonable regression model .