基于人格特质和机器学习分类算法的建筑工人不安全行为识别

获取建筑工人个性特征是实现其不安全行为精准化、个性化干预管理的重要前提,而人格特质是分析建筑工人个性特征的重要依据。本研究以292名一线建筑工人为研究对象,通过问卷调研和深度访谈探究人格特质与不安全行为之间的映射关系,基于大五人格生成不安全行为偏好,利用机器学习分类算法实现建筑工人的不安全行为识别。研究表明：高外倾性、中神经质、中宜人性、低责任心、低开放性映射习惯偏差型不安全行为；中外倾性、低神经质、低宜人性、低责任心、高开放性映射程序偏差型不安全行为；中外倾性、高神经质、中宜人性、高责任心、中开放性映射感知偏差型不安全行为；中外倾性、中神经质、中宜人性、中责任心、中开放性映射技能偏差型不安全行为。同时通过比选CART、RF、AdaBoost和GBDT四种分类算法模型的评估指标,结果发现GBDT算法的不安全行为预测性能最优。

Research on the identification of unsafe behavior of construction workers based on personality traits and machine learning classification algorithm

Obtaining the personality characteristics of construction workers is an important prerequisite for realizing the precise and individualized intervention management of their unsafe behaviors, and personality traits are an important basis for analyzing the personality characteristics of construction workers.This study took 292 front-line construction workers as the research objects, explored the mapping relationship between personality traits and unsafe behaviors through questionnaires and in-depth interviews, generated unsafe behavior preferences based on the Big Five, and used machine learning classification algorithms to achieve Unsafe behavior identification.Research shows: high extroversion, moderate neuroticism, moderate agreeableness, low conscientiousness, low openness mapping habit-biased unsafe behavior; moderate extroversion, low neuroticism, low agreeableness, low conscientiousness, high openness mapping program Deviant unsafe behavior; moderate extroversion, high neuroticism, moderate agreeableness, high conscientiousness, moderate openness Map perception of deviant unsafe behavior; moderate extroversion, moderate neuroticism, moderate agreeableness, moderate conscientiousness, moderate openness Mapping skill-biased unsafe behavior.At the same time, by comparing the evaluation indicators of the four classification algorithm models of CART, RF, AdaBoost and GBDT, it is found that the GBDT algorithm has the best performance in predicting unsafe behavior.