Yarnold P R, Soltysik R C, McCormick W C, Burns R, Lin E H, Bush T, Martin G J
Department of Medicine, Northwestern University Medical School, Chicago, Illinois 60611, USA.
J Gen Intern Med. 1995 Nov;10(11):601-6. doi: 10.1007/BF02602743.
To illustrate the use of multivariable optimal discriminant analysis (MultiODA).
Data from four previously published studies were reanalyzed using MultiODA. The original analysis was Fisher's linear discriminant analysis (FLDA) for two studies and logistic regression analysis (LRA) for two studies.
In Study 1, FLDA achieved an overall percentage accuracy in classification (PAC) for the training sample of 69.9%, compared with 73.5% for MultiODA. In Study 2, the LRA model required three attributes to achieve a 76.1% overall PAC for the training sample and a 79.4% overall PAC for the hold-out sample. Using only two attributes, the MultiODA model achieved similar values. In Study 3, the FLDA model achieved an overall PAC of 82.5%, compared with 87.5% for the MultiODA model. In Study 4, MultiODA identified a two-attribute model that achieved a 93.3% overall training PAC, when an LRA model could not be developed.
MultiODA identified: a superior training model (Study 1); a more parsimonious model that achieved superior overall training and identical hold-out PAC (Study 2); a model that achieved a higher hold-out PAC (Study 3); and a two-attribute model that achieved a relatively high PAC when a multivariable LRA model could not be obtained (Study 4). These findings suggest that MultiODA has the potential to improve the accuracy of predictions made in general internal medicine research.
阐述多变量最优判别分析（MultiODA）的应用。
使用MultiODA对四项先前发表的研究数据进行重新分析。其中两项研究的原始分析方法是费舍尔线性判别分析（FLDA），另外两项研究的原始分析方法是逻辑回归分析（LRA）。
在研究1中，FLDA对训练样本的分类总体准确率（PAC）为69.9%，而MultiODA为73.5%。在研究2中，LRA模型需要三个属性才能使训练样本的总体PAC达到76.1%，留出样本的总体PAC达到79.4%。仅使用两个属性时，MultiODA模型就能达到类似的值。在研究3中，FLDA模型的总体PAC为82.5%，而MultiODA模型为87.5%。在研究4中，当无法建立LRA模型时，MultiODA识别出一个双属性模型，其训练样本的总体PAC达到了93.3%。
MultiODA识别出：一个更优的训练模型（研究1）；一个更简洁的模型，该模型在总体训练中表现更优且留出样本的PAC相同（研究2）；一个留出样本PAC更高的模型（研究3）；以及一个在无法获得多变量LRA模型时能达到相对较高PAC的双属性模型（研究4）。这些发现表明，MultiODA有潜力提高普通内科研究中预测的准确性。
