glm 模型的预测概率是 0 还是 1?

Are predicted probabilities from glm models probabilities of 0 or 1?

我的响应变量,status 有两个值,1 表示活着,0 表示死了。

我已经建立了一个这样的模型 model<- glm(status ~., train_data, family='binomial')。我使用 predict(model, test_data, type = 'response'),它给出了一个预测概率向量,如下所示:

0.02  0.04  0.1

这些概率是有人活着(即 status == 1)还是有人死了(即 status == 0)?

我很确定这是一个人活着的概率,但是总是这样吗?有没有办法直接在 predict() 函数中指定它?

来自 ?binomial:

For the ‘binomial’ and ‘quasibinomial’ families the response can be specified in one of three ways:

  1. As a factor: ‘success’ is interpreted as the factor not having the first level (and hence usually of having the second level).
  1. As a numerical vector with values between ‘0’ and ‘1’, interpreted as the proportion of successful cases (with the total number of cases given by the ‘weights’).
  1. As a two-column integer matrix: the first column gives the number of successes and the second the number of failures.

如果 status 是数值为 0 或 1 的数值,则假定“案例总数”为 1(即,每个观察结果是失败 (0) 或成功 (1)单身人士)。 (概率是 总是“1 的概率”,即 0 总是表示“失败”,1 总是表示“成功”。)

据我所知,在 predict() 中无法更改此设置:如果您想翻转概率,则需要使用 1-status 而不是 status您的响应变量。