分类精确性指数Entropy在潜剖面分析中的表现:一项蒙特卡罗模拟研究

发布时间：2018-03-27 15:46

本文选题：潜剖面分析　切入点：分类精确性　出处：《心理学报》2017年11期

【摘要】：本研究通过蒙特卡洛模拟考查了分类精确性指数Entropy及其变式受样本量、潜类别数目、类别距离和指标个数及其组合的影响情况。研究结果表明:(1)尽管Entropy值与分类精确性高相关,但其值随类别数、样本量和指标数的变化而变化,很难确定唯一的临界值;(2)其他条件不变的情况下,样本量越大,Entropy的值越小,分类精确性越差;(3)类别距离对分类精确性的影响具有跨样本量和跨类别数的一致性;(4)小样本(N=50~100)的情况下,指标数越多,Entropy的结果越好;(5)在各种条件下Entropy对分类错误率比其它变式更灵敏。
[Abstract]:In this study, Monte Carlo simulations were used to investigate the effects of the classification accuracy index (Entropy) and its variations on the sample size, the number of latent categories, the distance and the number of indexes and their combinations. The results show that the Entropy value is highly correlated with the accuracy of the classification. However, its value varies with the number of categories, samples and index numbers, and it is difficult to determine the only critical value.) if other conditions are not changed, the larger the sample size is, the smaller the value of Entropy is. The worse the classification accuracy is, the worse the effect of classification distance on classification accuracy is. Entropy is more sensitive to classification error rate than other variants under various conditions.
【作者单位】：广州大学心理系;广州大学心理测量与潜变量建模研究中心;广东省未成年人心理健康与教育认知神经科学实验室;中国政法大学社会学院;
【基金】：国家自然科学基金(31400904) 广州大学“创新强校工程”青年创新人才类项目(2014WQNCX069);广州大学青年拔尖人才培养项目(BJ201715)
【分类号】：B841

【相似文献】