口医学 例:在医学诊断中,一个病人肺部有阴影, 医生要判断他患的是肺结核、肺部良性肿鸨 还是肺癌? 肺结核病人、肺部良性肿瘤病人、肺癌病人 组成三个总体,病人来自其中一个总体,可 通过病人的指标(阴影大小、边缘是否光滑 等)用判别分析判断他来自哪个总体(即判 断他患的什么病?) 2021/1/21 16
2021/1/21 16 cxt 医学： 例：在医学诊断中，一个病人肺部有阴影， 医生要判断他患的是肺结核、肺部良性肿瘤 还是肺癌？ 肺结核病人、肺部良性肿瘤病人、肺癌病人 组成三个总体，病人来自其中一个总体，可 通过病人的指标（阴影大小、边缘是否光滑 等）用判别分析判断他来自哪个总体（即判 断他患的什么病？）
a Discriminant Analysis Procedure: Obtain a random sample of objects from each class( these are objects s Those membership is known)This is known as the training or learning sample Submit the training sample to a Discriminant Analysis and obtain a set of discriminant functions. These functions are used implicitly by for example spss or sas. so you do not need to see or know them the information on these functions is stored in a dataset that is created within the program The same procedure allows a true validation of the classification functions by using a file that contains objects of known membership to be classified using only the information on the variables and the classification functions developed with the training"or learning sam nple 2021/1/21 cXt
2021/1/21 17 cxt Discriminant Analysis Procedure: • Obtain a random sample of objects from each class (these are objects whose membership is known) This is known as the “training” or “learning” sample. • Submit the training sample to a Discriminant Analysis and obtain a set of discriminant functions. These functions are used implicitly by for example SPSS or SAS, so you do not need to see or know them. The information on these functions is stored in a dataset that is created within the program. • The same procedure allows a true validation of the classification functions by using a file that contains objects of known membership to be classified using only the information on the variables and the classification functions developed with the “training” or “learning” sample
a the difference( or correlation between discriminant analysis and cluster analysis discriminant analysis classify new sample, in which we initially know how many distinct groups exist and we have data that are known to come from each of these distinct group. cluster analysis involves techniques that produce classifications from data that are initially unclassified 2021/1/21 18
2021/1/21 18 cxt the difference (or correlation) between discriminant analysis and cluster analysis : discriminant analysis classify new sample, in which we initially know how many distinct groups exist and we have data that are known to come from each of these distinct group. cluster analysis involves techniques that produce classifications from data that are initially unclassified
令判别分析与聚类分析的比较: (1)判别分析是在已知研究对象分成若干类型并已取得各种类 型的一批已知样本的观测数据,在些基础上根据某些准则建 立判别式,然后对未知类型的样本进行判别分类。 (2)聚类分析则是对研究对象的类型未知的情况下,对其进行分 类的方法 (3)判别分析和聚类分析往往联合使用。当总体分类不清楚时 先用聚类分析对一批样本进行分类,再用判别分析构建判别 式对新样本进行判别。 2021/1/21 19 cXt
2021/1/21 19 cxt ❖ 判别分析与聚类分析的比较： (1)判别分析是在已知研究对象分成若干类型并已取得各种类 型的一批已知样本的观测数据，在此基础上根据某些准则建 立判别式，然后对未知类型的样本进行判别分类。 (2)聚类分析则是对研究对象的类型未知的情况下，对其进行分 类的方法。 (3)判别分析和聚类分析往往联合使用。当总体分类不清楚时， 先用聚类分析对一批样本进行分类，再用判别分析构建判别 式对新样本进行判别
a Compared discriminant anal ysis with regression analysis Discriminant analysis is similar to regression analysis except that the dependent variable is categorical rather than continuous In regression analysis, we want to be able to predict the value of a variable of interest based on a set of predictor variables In discriminant analysis, we want to be able to predict class membership of an individual observation based on a set of predictor variables. 2021/1/21 20 cXt
2021/1/21 20 cxt Compared Discriminant analysis with regression analysis： Discriminant analysis is similar to regression analysis except that the dependent variable is categorical rather than continuous. In regression analysis, we want to be able to predict the value of a variable of interest based on a set of predictor variables. In discriminant analysis, we want to be able to predict class membership of an individual observation based on a set of predictor variables