摘要
目的采用随机森林算法分析体检人群肾结石的影响因素。方法自体检人群中选取955例肾结石患者和1 670例未患肾结石者,收集各项生化指标,先利用随机森林方法进行降维,再用传统的Logistic回归对降维后的变量进行分析。结果经随机森林算法筛出8个重要性得分最高且错误率最低的变量纳入经典Logistic回归模型进行分析,最终进入Logistic回归模型的变量有性别、年龄、体质指数、收缩压、低密度脂蛋白、总胆红素。结论肾结石的发病与性别、年龄及人体多项生化指标有关。
Objective To investigate the risk factors of nephrolithiasis with Random Forest model. Methods Randomly selected patients with kidney stones 955 cases and 1670 cases not suffering from kidney stones were collected physiological and biochemical indices of each group. Then we analyzed the risk factors for kidney stones using Random Forest. Results Eight important variables,whose average importance scores were highest and whose error rates were lowest, were selected by Random Forest method. Then the logistic regression was conducted, including the variables of sex, age, BMI, SBP, LDL, TBIL. Conclusion The increase of age, BMI,SBP, and LDL are the risk factors for renal calculus, but the increase of TBIL, sex showed inhibition against renal calculus.
出处
《现代预防医学》
CAS
北大核心
2016年第1期1-3,7,共4页
Modern Preventive Medicine
基金
社会发展项目(BE2011647)
关键词
肾结石
随机森林
影响因素
Nephrolithiasis
Random Forest
Factors
作者简介
李苹（1989-），女，硕士，研究方向：流行病与卫生统计学
通讯作者：黄水平，E—mail：hsp@xzmc．edu．cn
