论文部分内容阅读
目的:探究将统计学习方法应用于心理测验所得的大量数据进行学习分析的可行性,并基于探究结果对飞行职业的人格特征进行进一步探索,为飞行人员的选拔及评估提供新的思路。方法:从某航空公司随机抽取1020名男性被试,其中飞行人员510名,非飞行人员510名,采用卡特尔16项人格测试对其进行测验,施测后对得到的16项因子分采用支持向量机就随机划分的训练组和测试组进行学习,分析学习结果。结果:挑选出4项因子作为分类的特征因子,基于线性支持向量机构建的分类器在交叉验证下的平均正确率为64%。结论:采用SVM构建的分类器具有一定的可靠性和有效性。
OBJECTIVE: To explore the feasibility of applying statistical learning methods to a large number of data obtained from psychological tests and to further explore the personality characteristics of flight occupations based on the findings, so as to provide new ideas for the selection and evaluation of flight crews. Methods: A total of 1020 male subjects were selected from an airline company, including 510 pilots and 510 non-pilots. The test was conducted using 16 personality tests of cartels. After the test, the 16 factors were used to support Vector machine training groups and test groups randomly divided learning, analysis of learning outcomes. Results: Four factors were selected as feature factors of classification. The average accuracy of the classifier based on linear support vector machine was 64% under cross-validation. Conclusion: The classifier constructed by SVM has certain reliability and validity.