首页|基于Logistic回归模型的噪声作业工人高频听力损伤影响因素分析

基于Logistic回归模型的噪声作业工人高频听力损伤影响因素分析

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
为确定造成听力损伤的关键原因,本研究采用Logistic回归模型,探究噪声环境下作业工人高频听力损伤的关键因素.通过对比噪声环境与非噪声环境工人的听力、血压等数据,结合随机森林算法评估影响因素的重要性.最后选取同时满足重要性评分最高、袋外数据袋外预测残差均方最小的影响因素作为自变量,以是否为高频听力损伤为因变量,通过Logistic回归模型分析噪声作业工人高频听力损伤的影响因素.试验结果表明,噪声对噪声作业人员身体健康具有负面影响,高频听力损伤发生率达到45.71%.经常或总是采取抗噪防护措施可降低高频听力损伤风险,而长时间噪声暴露、高强度噪声、高龄、特定工种和长时间耳机使用等因素增加了高频听力损伤的风险.
Analysis of Factors Affecting High-frequency Hearing Loss in Noise Workers Based on Logistic Regression Model
To identify the key causes of hearing impairment,a logistic regression model is used to explore the key factors of high-frequency hearing impairment of working workers in a noisy environment.By comparing the hearing,blood pressure and other data of workers under noise environment with those data under non-noise environment,combined with random forest algorithm,the importance of affecting factors is assessed.The factors with the most importance and the minimum prediction mean square error are selected as independent variables,and high frequency hearing impairment as the dependent variable,logistic regression model is used to analyze the factors leading to high frequency hearing impairment.Experimental results show that the noise has significant negative impact on worker health,causing high frequency hearing damage incidence rate of up to 45.71%.Frequently or consistently taking noise protection measures can reduce the risk of high frequency hearing damage,while the risk of high frequency hearing damage will increase due to long time noise exposure,high intensity noise,old age,specific type of work and long time use of headset.

Logistic regressionnoise-exposed workhigh frequency hearing lossinfluencing factorsrandom forestresidual mean square

仇欢、张家祥、李旭东

展开 >

安徽医科大学公共卫生学院,安徽 合肥 230032

广东省职业病防治院/广东省职业健康工程技术研究中心,广东 广州 510300

Logistic回归 噪声作业 高频听力损伤 影响因素 随机森林 残差均方

广东省自然科学基金国家重点研发计划项目西部绿色建筑国家重点实验室项目广东省职业健康工程技术研究中心建设项目广州市基础研究计划项目2023年度合肥综合性国家科学中心大健康研究院职业医学与健康联合研究中心科研项目

2023A15150113292022YFC2503205LSKF202313D:2019A069202002030396OMH-2023-19

2024

系统仿真技术
同济大学

系统仿真技术

CSTPCD
影响因子：0.271
ISSN：1673-1964
年,卷(期)：2024.20(2)
  • 9