Logistic regression: Meaning of beta values

2 次查看（过去 30 天）
Pauli
Pauli 2019-5-7
Hello,
I have a dummy question about how to interpret the beta values of mnrfit. I plotted the logistic regression, see below, (one dependent variable) and have two beta values in my stats file. The top one is -0.5094, the second one is 0.0633. In the documentation of mnrfit it says that the first is the intercept and the second is the slope. Now, do I understand it correctly that those are log odds?
The fit contains predicted probabilities, so I have to transfer the beta values also. If I try to do so (exp(beta)/(1+exp(beta)), values I wouldn't expect are returned. I would expect sth like 0.64 for the intercept and a negative value for the slope.
Can anybody help me with this?
Thank you.
ASK.png

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Linear Regression 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by